Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Human Hand Motion Analysis during Different Eating Activities

Human Hand Motion Analysis during Different Eating Activities Hindawi Applied Bionics and Biomechanics Volume 2018, Article ID 8567648, 12 pages https://doi.org/10.1155/2018/8567648 Research Article Zakia Hussain , Norsinnira Zainul Azlan , and Arif Zuhairi bin Yusof Department of Mechatronics Engineering, International Islamic University Malaysia, Gombak, 53100 Kuala Lumpur, Malaysia Correspondence should be addressed to Norsinnira Zainul Azlan; sinnira@iium.edu.my Received 6 October 2017; Revised 24 November 2017; Accepted 20 December 2017; Published 4 February 2018 Academic Editor: Laurence Cheze Copyright © 2018 Zakia Hussain et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The focus of this research is to analyse both human hand motion and force, during eating, with respect to differing food characteristics and cutlery (including a fork and a spoon). A glove consisting of bend and force sensors has been used to capture the motion and contact force exerted by fingers during different eating activities. The Pearson correlation coefficient has been used to show that a significant linear relationship exists between the bending motion of the fingers and the forces exerted during eating. Analysis of variance (ANOVA) and independent samples t-tests are performed to establish whether the motion and force exerted by the fingers while eating is influenced by the different food characteristics and cutlery. The middle finger motion showed the least positive correlation with index fingertip and thumb-tip force, irrespective of the food characteristics and cutlery used. The ANOVA and t-test results revealed that bending motion of the index finger and thumb varies with respect to differing food characteristics and the type of cutlery used (fork/spoon), whereas the bending motion of the middle finger remains unaffected. Additionally, the contact forces exerted by the thumb tip and index fingertip remain unaffected with respect to differing food types and cutlery used. 1. Introduction during eating, an in-depth knowledge of hand motion during eating is vital. Upper limb disability is one of the major adversities faced by Hand motion during eating is highly dexterous and is poststroke patients. The resulting loss of mobility in these subject to the type of food ingested and the type of cutlery patients reduces their ability to perform normal activities used. Analysing hand motion can be complicated due to its of daily living (ADL), preventing them from leading a nor- highly articulate nature. A human hand consists of 27 bones mal life and hence reducing their quality of life. These and 35 muscles, of which 17 are intrinsic muscles (located in patients are highly dependent on their caregivers (usually a the palm) and 18 are extrinsic muscles (located in the fore- arm). With roughly 30 degrees of freedom (DOFs), this com- spouse or friend) who perform most of their basic ADL, such as eating, bathing, and grooming, which gradually has plex structure can perform intricate tasks, which require a negative impact on the mental and physical state of the dexterity. During the past few years, hand motion analysis caregiver [1–6]. has gained the attention of the researchers working in the Eating is one of the fundamental activities of survival for field of rehabilitation, human-computer interaction (HCI), all living beings. Dysphagia and other eating difficulties are and robotics. also common among poststroke patients which can lead to Hand motion analysis enables researchers to gather data complications, such as malnutrition, dehydration, suffoca- such as the force applied by the fingers, different joint angles tion, and eventually death [7–10]. Over the past decade, of the hand, and velocity, while performing different activi- numerous robotic rehabilitation systems have been devel- ties. Analysing the motion and force of the hand during var- oped to assist impaired patients regain their hand functions. ious eating activities can help in formulating a model, which Such robotic systems must have the capability to replicate in turn can be useful in developing a rehabilitation robot for human hand function during any ADL. To develop a rehabil- assisted eating. Several studies have been conducted to ana- itation system meant specifically to regain the hand function lyse the motion of the hand and upper limb while performing 2 Applied Bionics and Biomechanics discussion of the data analysis results in the previous section, different daily activities of living. Ju and Liu [11], Gopura et al. [12], and Tang et al. [13] have successfully analysed and lastly, the conclusion is drawn in Section 5. and classified different human hand motions while perform- ing basic daily activities, such as hair combing and recogniz- 2. Experimental Method ing multiple hand gestures, using electromyography (EMG). In EMG analysis, tiny electrodes, when placed on human 2.1. Experimental Setup. A prototype glove has been used skin, detect and record the electrical signals transmitted by to analyse the motion of hand during eating (Figure 1). the motor neurons responsible for activating muscle contrac- The glove for hand has been designed as an instrument tion. Ju and Liu [11] used a framework of multiple sensor to measure the angle of the index finger, middle finger, and integration of CyberGlove, Finger TPS pressure sensors, thumb. The glove is developed with three flexible bend sen- and Trigno wireless EMG sensors to capture hand gestures, sors (Spectra Symbol, 4.5 inches) for measuring the angles contact forces, and muscle contraction signals from various of the index finger, middle finger, and thumb (Figure 2). hand motions, while performing 10 basic grasping activities, These bend sensors act as variable resistors which, when such as holding and lifting a dumbbell and opening and clos- flexed, increase the resistance across the sensor. Force sensors ing a pen box, using five fingers. (FlexiForce™, A201) are attached to the finger tip of the index Cabibihan et al. [14] explored the human patting ges- and thumb to measure the force exerted by the thumb and ture for analysing the amount of force applied to regions index finger, during eating process, since only the index fin- of the hand and the angular motion of finger joints so as ger and thumb are involved in holding the spoon/fork during to incorporate them into a humanoid robot, in order to any eating activity. imitate this gesture. Similarly, the kinematics and dynamics The data from the glove is recorded using MATLAB 2015 of the human arm, during 24 daily activities (such as eating through serial communication with Arduino. (Figure 3) using a spoon and a fork, drinking with a cup, and washing demonstrates the hardware setup of the bend sensors and the face) were studied by Rosen et al. [15] to develop a 7- the force sensors. DOF powered exoskeleton for the upper limb. Ah et al. [16] performed human hand motion analysis while turning 2.2. Data Acquisition. Six healthy, right-handed subjects a door knob. including three males and three females, age ranging from Aprile et al. [17] dedicated an entire study to analyse the 24 to 30 years and an average weight of 65 kgs, volunteered upper limb motion in stroke patients while performing a for this study. Five eating activities were performed, to ana- drinking task, which included reaching for the glass, bringing lyse the hand motion, while using different eating cutlery it to the mouth, and putting it back on the table. Adnan et al. (spoon and fork) and food types (including solids and liq- [18] developed a low-cost DataGlove using a flexible bend uids). The type of food involved in the eating activities sensor to recognize various human finger activities. In addi- included cooked rice, milk cereal, salad with chunks of vege- tion, the analytical mathematical model and analysis of vari- tables, noodles, and a clear soup broth. A plastic spoon and a ance (ANOVA) was established to predict the force induced steel fork were used during the activity. Each activity was per- at the flexible force sensor by the human finger using the low- formed three times by each participant, with each trial lasting cost DataGlove [19]. seven seconds and while sitting on a chair with food on the Some previous work on hand analysis is summarized in table. The five activities performed were as follows: Table 1. Despite many studies on hand motion (Table 1), to (1) Eating rice (solid) with a spoon the best of our knowledge, there has not been a study dedi- cated to the analysis of hand motion while eating different (2) Eating soup broth (liquid) with a spoon types of food and using different cutlery. It is important to consider the food characteristics and the amount of force (3) Eating cereal with milk (mixture of solid and liquid) exerted by the hand during eating to enhance the develop- using a spoon ment of robotic rehabilitation systems for this activity. (4) Eating vegetable salad (solid) using a fork Therefore, this paper presents human hand motion anal- ysis, focusing on the thumb, index finger, and middle finger (5) Eating noodles (solid) using a fork during eating. The motion of these three fingers and the force they exert during eating is studied with respect to the type of Subjects were trained before performing the activities on food (liquid, solid) and the cutlery used. An experiment has how to grasp the cutlery and eat using it, while wearing the been conducted involving five different food types and using glove. For eating noodles, subjects were asked to roll over two different types of cutlery (fork and spoon) to study their the noodles on the fork and then eat. Each trial of the eating effect on hand motion. ANOVA and t-test analysis has been activities performed consisted of four main events (Figure 4). conducted to study the influence of these factors on the finger The first event in each eating activity was the origin or start- motion and force during eating. The paper is organized as ing point, which occurred when the subject kept the glove follows: Section 2 presents the method of experimentation rested horizontally on the table and bend sensors with almost employed which is subdivided into two subsections: Experi- no bend. The second event called event A occurred, when the mental Setup and Data Acquisition. Section 3 presents the subject holding a spoon or fork digs in to the food and brings data analysis and results obtained while eating different types it towards the mouth to eat. The third event known as event B of food and using different cutlery. Section 4 presents the occurred, when the subject during eating maintains the grip Applied Bionics and Biomechanics 3 Table 1: Highlights of the previous contributions to human motion analysis. Data Number Authors Objective Focus of study acquisition Results/findings Activity method Strong correlations between To correlate the muscle Ten in-hand Human hand motion muscle signals, contact forces, signals with contact forces EMG sensor, manipulations Ju and Liu analysis with and finger trajectories. 1 and finger trajectories & force sensor & like holding & [11] multisensory Fuzzy Gaussian mixture DataGlove lifting a motion recognition using information models (FGMMs) used for muscle signals dumbbell motion recognition To analyse upper-limb EMG Relationships between the Basic motions muscle activities during Human upper-limb electrodes, upper limb motions & and the Gopura basic upper-limb motion, muscle activities 2 VICON activity levels of main selected daily et al. [12] to design power-assist during daily upper- motion muscles have been activities of robotic exoskeleton limb motions capture system established upper-limb systems Hand motion Experimental results showed To classify multiple hand Tang et al. classification using a that the success rate for the 11 hand 3 gestures using three sEMG sensors [13] multichannel surface identification of all the 11 gestures different methods sEMG sensor gestures is significantly high To analyse the gesture, the amount of force applied Human patting gesture CyberGlove II The sensitive regions on the Cabibihan Human 4 on regions of the hand, analysis for robotic FingerTPS hand while performing pat et al. [14] patting gesture and the angular motion of social touching sensors have been identified finger joints VICON The results indicated that the To study the kinematics The human arm motion various joints’ kinematics and Rosen and the dynamics of the kinematics and 5 capture system dynamics change 24 ADL human arm during daily dynamics during daily et al. [15] &reflective significantly based on the activities activities markers nature of the task To evaluate motor control 3D kinematic motion VICON abilities between the analysis of door motion Comparisons have been Ah et al. Door handling 6 groups of people with mild handling task in capture system drawn between healthy, mild [16] task and moderate arm people with mild and &reflective & moderate stroke patients impairments moderate stroke markers To analyse, using motion Comparisons have been Smart motion analysis, the qualitative Kinematic analysis of drawn between stroke & Aprile capture 7 and quantitative upper the upper limb motor healthy control group while Drinking task et al. [17] optoelectronic limb motor strategies in strategies in stroke reaching out for the glass to system stroke patients drink Low-cost Measurement of the Sign language To develop a low-cost DataGlove by flexible bending force The DataGlove developed translation Adnan DataGlove, able to using the 8 of the index and can measure several human (letters A, B, C, et al. [18] recognize the different flexible middle fingers for degrees of freedom (DoFs) D, F & K and finger activities bending virtual interaction number 8) sensor Accurate An analytical mathematical measurement of force model and ANOVA has been To find the correlations Flexiforce Any finger Adnan by the force sensor for established to predict the 9 between the forces of pressure gripping et al. [19] intermediate and force induced at the flexible finger phalanges sensors activity proximal phalanges of force sensor and the human index finger finger of low-cost DataGlove on the cutlery. The final event is known as event C, when the 3. Data Analysis and Results subject releases the cutlery after eating and goes back again to the origin, that is, the subject after finishing the eating 3.1. Bending Finger Motion Trajectories. The motion trajecto- brought his/her hand back to rest on the table. Throughout ries captured by the bend sensors for the thumb, index finger, the experiment, subjects were asked to keep their elbows and middle finger during five different eating activities are rested on the table. shown (Figure 5) with the four important events identified 4 Applied Bionics and Biomechanics Force sensors Bending sensors (a) (b) Figure 1: (a) The location of bend sensors on the thumb, middle finger, and index finger. (b) The location of force sensors on the thumb and the index finger. (a) (b) (c) Figure 2: (a) The index finger angle, (b) the middle finger angle, and (c) the thumb angle, measured by the bend sensors. Data from glove (bend sensors & Arduino MATLAB Computer force sensors) Figure 3: Hardware setup of the bend and force sensors for hand motion analysis. Origin Event A (a) (b) Event C Event B (c) (d) Figure 4: Four main events identified during each eating activity: (a) origin, (b) event A, (c) event B, (d) Event C. Applied Bionics and Biomechanics 5 umb motion trajectory Index finger motion trajectory Event A Event A 70 Event C Event C Origin Event B Origin Event B Origin Origin 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 Time (s) Time (s) Rice Veg Rice Veg Cereal Soup Cereal Soup Noodle Noodle (a) (b) Middle finger motion trajectory Event A Event C Event B Origin Origin 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 Time (s) Rice Veg Cereal Soup Noodle (c) Figure 5: (a) Thumb motion trajectories, (b) index finger motion trajectories, and (c) middle finger motion trajectories obtained from the bend sensor for five different eating activities. on the trajectory. During all five eating activities (rice, cereal, maximum averaged angle of 121.8 ; for noodle eating activ- and soup with a spoon, noodle and vegetable using a fork), ity, the ROM is from a minimum averaged angle 17 to a the averaged range of motion (ROM) for the thumb ranged maximum averaged angle of 115 ; for cereal with milk ° ° from a minimum of 19.5 activity using a spoon, the ROM is from a minimum aver- to a maximum of 59.1 (referring ° ° ° to Figure 2). The origin is around 19 , although the subjects aged angle of 17 to a maximum averaged angle of 111 ; kept their hands horizontally at rest on the table; this can for eating vegetables using a fork, the ROM is from a mini- ° ° be due to the sensor fatigue while doing the activities repeat- mum averaged angle of 17 to a maximum of 103 ; and for edly (Figure 5(a)). the soup broth eating activity using a spoon, the ROM is The ROM of the index finger during eating rice with a from a minimum of 17 to a maximum averaged angle of ° ° spoon is from a minimum averaged angle of 7 to a maxi- 120.6 (Figure 5(c)). mum averaged angle of 90 ; for the noodle eating activity, From the graphs (Figures 5(a)–5(c)), it can be observed the ROM is from a minimum averaged angle of 7 to a max- that as event A starts (at around 0.5 seconds), the bending imum averaged angle of 93 ; for cereal with milk activity angles of the fingers start increasing to grip the cutlery and reach a maximum value, while bringing the food to the using a spoon, the ROM is from a minimum averaged angle ° ° of 7 to a maximum averaged angle of 75 ; for vegetable eat- mouth. During event B (lasting around 3 seconds), the ing activity using a fork, the ROM is from a minimum aver- magnitude of the angles remains steady, while the subject ° ° aged angle of 9 to a maximum averaged angle of 83 ; and for maintains the grip on the cutlery during eating. The magni- the soup broth eating activity using a spoon, the ROM is from tude of the bending angles starts decreasing during event C ° ° a minimum of 7 to maximum averaged angle of 72 . The ori- (lasting around 2 seconds), when the subject releases the gin in all activities is around 7 (Figure 5(b)). grip off the cutlery by putting it back into the dish and pro- The ROM of the middle finger whist eating rice with ceeding towards the origin, when the bending angles again ° ° a spoon is from a minimum averaged angle of 18 to a settle at around 20 . Average angles (º) Average angles (º) Average angles (º) 6 Applied Bionics and Biomechanics Force exerted by index fingertip Force exerted by thumb tip 3.00E + 00 2.50E + 00 2.00E + 00 3 1.50E + 00 1.00E + 00 5.00E − 01 0.00E + 00 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Time (s) Time (s) Trial 1 Trial 1 Trial 2 Trial 2 Trial 3 Trial 3 (a) (b) Figure 6: (a) Three trials of thumb-tip force and (b) index fingertip force captured by the force sensor during vegetable eating activity. Force exerted by thumb tip Force exerted by index fingertip Event C 2.5 Event A Event A Event C 2.5 1.5 1.5 Origin Origin Event B Event B Origin 0.5 0.5 Origin 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 Time (s) Time (s) Rice Veg Rice Veg Cereal Soup Cereal Soup Noodle Noodle (a) (b) Figure 7: (a) Thumb-tip force and (b) index fingertip force recorded by the force sensor during five different eating activities. 3.2. Contact Force Trajectories. The force sensors attached subject is putting the spoon/fork back into the dish, eventu- to the prototype glove measured the force exerted by the ally, coming back to the origin, when the subject rests his/ thumb tip and the index fingertip, while performing five her hand on table again. During the noodle eating activity different eating activities, since only the thumb and index using a fork, the thumb tip exerts a maximum average force finger are involved in holding the spoon or fork while eating. of 2.7 N, which is the highest of all other eating activities. To check the repeatability of the force sensors used, a subject Since during the noodle eating activity, the subjects were performing three trials of the vegetable eating activity is asked to roll over the noodles on the fork; hence, the force shown in Figure 6. trajectory shows some minor fluctuations and is longer than From Figure 6, it can be observed that the force sensor other eating activities. For rice eating activity using a spoon, a measurements, attached to the index fingertip and the thumb maximum average force of 2.4 N; for cereal with milk eating tip, demonstrate quite consistent results. activity using a spoon, a maximum average force of 2.2 N; (Figures 7(a) and 7(b)) demonstrate the force exerted by for vegetable eating activity using a fork, a maximum average the thumb tip and the index fingertip with the four main force of 2.4 N; and for soup broth eating activity using a events identified on the graphs. During the origin, the hand spoon, a maximum average force of 2.0 N is exerted by the is lying horizontally on the table, and as such, no force is thumb tip (Figure 7(a)). During the cereal with milk eating activity using a spoon, exerted by the thumb tip/index fingertip, but as event A starts and the subject grips a spoon or fork between the thumb and the index fingertip exerts a maximum average force of 2.4 N, index finger, the magnitude of force increases and during which is the highest of all other eating activities. For rice eat- event C and a maximum force is reached, when the subject ing activity using a spoon, a maximum average force of 2.0 N; is digging into the food or trying to get the food on the spoon for noodle eating activity using a fork, maximum average force of 2.0 N; for vegetable eating activity using a fork, a or fork. The force then starts to decrease in event C, as the Force (n) Force (n) Force (n) Force (n) Applied Bionics and Biomechanics 7 Table 2: Averaged Pearson coefficient of bending finger angles and force exerted by fingers. Rice (spoon) Cereal & Milk (spoon) Soup (spoon) Vegetable (fork) Noodle (fork) Index Middle Thumb Index Middle Thumb Index Middle Thumb Index Middle Thumb Index Middle Thumb FINDEX 0.93 0.87 0.78 0.89 0.84 0.95 0.90 0.87 0.94 0.89 0.81 0.74 0.85 0.73 0.89 FTHUMB 0.90 0.84 0.81 0.89 0.83 0.94 0.90 0.89 0.94 0.90 0.83 0.80 0.86 0.72 0.88 Table 3: Analysis of variance summary table. maximum average force of 2.2 N; and for soup broth eating activity using a spoon, a maximum average of 2.1 N force is Sum of Mean Sig./p df exerted by the index fingertip (Figure 7(b)). squares square value Between 3.3. Correlations between Bending Angles and the Contact 893.33 4 223.33 0.023 groups Forces of Fingers. The bending finger angle data and the con- BENDTHMB Within tact force data captured by the sensors has been used to find 26855.55 345 77.84 groups the correlations between the bending angles of the thumb, Total 27748.88 349 index finger, and middle finger and the contact forces exerted by the thumb and index finger, during eating activities using Between 5765.47 4 1441.37 0.074 the Pearson product moment correlation coefficient (PPMC). groups Pearson correlation coefficient (r) measures the strength BENDINDX Within 230984.22 345 669.52 and direction of a linear relationship between two variables. groups The SPSS statistics software package has been used to per- Total 236749.69 349 form this analysis. It takes values ranging from +1 to −1; Between 3236.77 4 809.19 0.609 r =+1 implies a strong positive linear relationship between groups the variables, while r = 1 implies a strong negative linear rela- BENDMID Within tionship, and r =0 implies no linear relationship between the 413100.47 345 1197.39 groups variables. Equation (1) gives the formula for computing the Total 416337.24 349 Pearson correlation coefficient [20]. n〠 xy − 〠x 〠y groups. SPSS Statistics software package has been used to r = , 1 perform ANOVA in this study. In this study, ANOVA has 2 2 2 2 n 〠x − 〠x n 〠y − 〠y been used to determine whether the bending motion of the index finger (BENDINDX), middle finger (BENDMID), and thumb (BENDTHMB)differed based on different groups of where n = number of data pairs. food types (cereal, rice, vegetable, noodle, and soup) or not The averaged Pearson coefficients of different bending (Tables 3 and 4). finger angles and the forces exerted by the index finger and The df column in Table 3 means degrees of freedom, thumb, for all six subjects involved in the experiment, are which is the division of Sum of squares by Mean square values shown in (Table 2). in the ANOVA summary table. The Sum of squares is the The most significant coefficients are highlighted sum of Between groups and Within groups. The Sig. column (Table 2). From the results (Table 2), both the soup (liquid) denotes the p value which represents the probability of find- and cereal (solid and liquid) eating activities which are per- ing an effect equal to or greater than the one observed consid- formed using a spoon have similar results, where both index ering the null hypothesis to be true. The null hypothesis here fingertip force and the thumb-tip force have shown the stron- signifies that there is no significant difference in the bending gest positive linear relationship with the average bending angles of the thumb, index finger, and middle finger with motion of the thumb. Noodle eating activity has shown sim- respect to different types of food groups. ilar results with the soup and cereal activities. Moreover, dur- The lower the p value, the more likely the null hypothesis ing the vegetable and rice eating activities, the force exerted is rejected (preferably less than 0.05, while 0.10 is also by the thumb tip and the index fingertip have the strongest accepted but as a weak evidence). The p value thus provides positive correlation with the averaged index finger motion. a quantitative strength of evidence against the null hypothe- During all the eating activities, the middle finger motion sis stated [21, 22]. From (Table 3), it can be concluded that has the weakest linear relationship with the index fingertip for the average bending motion of the thumb and the index and thumb-tip force as compared to the bending motion of finger across the five groups of food (soup, rice, noodle, the index finger and thumb. cereal, and vegetable salad), there is a statistically significant 3.4. One-Way ANOVA of the Bending Angles of Fingers with difference at 5% and 10%, respectively, whereas for the aver- Respect to Different Types of Food. The one-way analysis of aged bending motion of the middle finger, there is no statis- tically significant difference, while eating different types of variance or ANOVA is a statistical comparison test used to determine whether there are any statistically significant dif- food (p > 10%). In simple words, it can be concluded from the ANOVA results that the bending motion of the thumb ferences between the means of two or more independent 8 Applied Bionics and Biomechanics Table 4: A least significant difference post hoc test using SPSS An LSD (least significant difference) post hoc test has software package. been carried out to distinguish eating precisely which type of food group (noodle, soup, rice, vegetables, and cereal) Dependent (I) (J) Mean difference Sig. the variances occur in the bending motions of the thumb variable foodtype foodtype (I − J) and the index finger (Table 4). To check for the vari- BENDTHMB ances, Mean difference (I-J) and Sig. (significance/p value) Rice −2.66 0.075 column of Table 4 is considered. The LSD results for the bending angles of the thumb from Table 4 can be sum- Veg 2.33 0.119 1 Cereal marised as follows: Noodle −0.60 0.687 Soup −0.61 0.684 (1) The bending angles of thumb during cereal eating Cereal 2.66 0.075 activity are smaller than its bending angles during 4.99 Veg 0.001 rice eating activity at a statistically significant differ- 2 Rice Noodle 2.06 0.168 ence of 10% (p =0 075, read row 1). Soup 2.06 0.169 (2) On the contrary, the bending angles of thumb during Cereal −2.33 0.119 vegetable, noodle, and soup eating activities show sta- −4.99 Rice 0.001 tistically no significant difference with the bending 3 Veg Noodle −2.93 0.050 motion of thumb during cereal eating activity. In Soup −2.94 0.050 other words, the bending motion of the thumb dur- ing vegetable, noodle, and soup eating activities does Cereal 0.60 0.687 not show any variance with respect to the cereal eat- Rice −2.06 0.168 4 Noodle ing activity. Veg 2.93 0.050 (3) The bending angles of thumb during rice eating Soup 0.00 0.998 activity are again greater than its bending angles dur- Cereal 0.61 0.684 ing the vegetable eating activity, at a statistically sig- Rice −2.06 0.169 5 Soup nificant difference of 1%, while the bending angles 2.94 Veg 0.050 of thumb during noodle and soup activities show sta- Noodle 0.00 0.998 tistically no significant difference with the bending BENDINDX motion of thumb during rice eating activity. Rice −3.29 0.452 (4) Similarly, the bending angles of thumb during Veg −3.33 0.447 vegetable eating activity are lesser than its bending 6 Cereal −9.97 Noodle 0.023 angles during noodle and soup eating activities at a statistically significant difference of 10% and 5%, Soup 1.96 0.655 respectively. Cereal 3.29 0.452 Veg −0.04 0.993 (5) Finally, the bending motion of the thumb during 7 Rice Noodle −6.68 0.127 noodle eating activity has statistically no significant difference with the bending motion of the thumb Soup 5.25 0.231 during soup eating activity. Cereal 3.33 0.447 Rice 0.04 0.993 Therefore, from the LSD post hoc results, it can be 8 Veg Noodle −6.65 0.129 concluded that the bending motion of the thumb during Soup 5.28 0.228 rice and vegetable eating activity is maximum as compared to other eating activities and the bending motion of the Cereal 9.97 0.023 thumb does not show much variance during noodle and Rice 6.68 0.127 9 Noodle soup eating activities. Veg 6.65 0.129 Similarly, the ANOVA results for bending angles of index 11.93 Soup 0.007 finger (Table 4) can be summarised as follows: Cereal −1.96 0.655 Rice −5.25 0.231 (1) The bending angles of index finger during cereal eat- 10 Soup Veg −5.28 0.228 ing activity are smaller than its bending angles during −11.93 the noodle eating activity, at a statistically significant Noodle 0.007 difference of 5%. The mean difference is significant at the 0.05 level. (2) Contrary, the bending angles of index finger during and the index finger is influenced by the type of food, rice, vegetable, and soup eating activity show statis- whereas the bending motion of the middle finger is not tically no significant difference with the bending motion of index finger during cereal eating activity. affected by type of food. Applied Bionics and Biomechanics 9 Table 5: ANOVA summary table. Table 6: Group statistics showing the mean and standard deviation (SD) for the bending motion data analysis. Sum of Mean Sig./p df squares square value Cutlery Mean Std. deviation Between Fork 41.39 28.29 0.30 4 0.08 0.892 BENDINDX groups Spoon 35.19 24.19 FINDX Within Fork 58.12 33.62 94.05 345 0.27 groups BENDMID Spoon 56.19 35.20 Total 94.35 349 Fork 33.02 8.81 Between BENDTHMB 3.54 4 0.88 0.273 Spoon 34.97 8.92 groups FTHMB Within 236.09 345 0.68 groups t-test is conducted instead of ANOVA because here, the fac- Total 239.63 349 tor (independent variable) which is the cutlery has only two groups (fork/spoon), but for conducting ANOVA, it must be more than two; hence, an independent samples t-test has (3) The bending angles of the index finger during rice eating activity show statistically no significant differ- been conducted. The results of the independent samples t-test are shown ence with the bending motion of the index finger dur- ing the vegetable, noodle, and soup eating activities. in Tables 6 and 7. To interpret the results from the t-test table (Table 7), the large column labelled Levene’s test for equality (4) During the vegetable eating activity, the bending of variances is checked first. This is a test that determines if motion of the index finger shows statistically no sig- the two conditions (a fork and a spoon) have about the same nificant difference with the bending motion of the or different amounts of variability between scores. Under this index finger during noodle and soup eating activities. column, the Sig. p value column is considered. This Sig. value determines which row to consider, either the Equal variances (5) Finally, during the noodle eating activity, the bending assumed or the Equal variances not assumed row. If the Sig. angles of the index finger are greater than its bending value is greater than 0.05, read from the top row, which angles, during soup eating activity at a statistically means that the variability in the two conditions is about the significant difference of 10%. same. That is, the scores in one condition (fork) do not vary Thus, from the LSD post hoc test results, it can be sum- much more than the scores in the second condition (spoon). marised that the bending motion of the index finger during Put scientifically, it means that the variability in the two con- noodle eating activity is highest as compared to the other four ditions is not significantly different and vice versa, if the Sig. eating activities. Additionally, the bending motion of the value is lesser or equal to 0.05. In the latter case, read from the index finger does not show any significant variance with bottom row, that the variability in the two conditions is not respect to soup, vegetable, and rice eating activities. the same. That is, the scores in one condition vary much more than the scores in the second condition. Scientifically, 3.5. One-Way ANOVA of the Forces Exerted by the Finger it means that the variability in the two conditions is signifi- Tips with Respect to Different Types of Food. The one-way cantly different. ANOVA technique has been again used to determine if there From Table 7, for the bending motion of the index finger exists a statistically significant difference in the averaged (BENDINDX), the p value is less than 0.05 (0.000), that is, forces exerted by the thumb tip (FTHMB) and the index fin- reading from the bottom row, which reveals that the variabil- gertip (FINDX), based on the different groups of food type ity in the two conditions (fork and spoon) is not the same. (Table 5). From the ANOVA results, it can be concluded that After finding the row to read (bottom row), now the results for both forces exerted by the thumb and the index finger, of t-test can be found in the column labelled t-test for equality there exists statistically no significant difference during vari- of mean by considering the Sig. (2-tailed) column under it. ous eating activities (p > 10%). As seen from the Sig. column This Sig. (2-tailed) column determines if the two conditions’ in Table 5, for both index finger and the thumb, the Sig. is means are statistically different. If the Sig. (2-tailed) is greater 0.892 (89.2%) and 0.273 (27.3%), respectively, which is far than 0.05, this means that there is no statistically significant greater than the desired Sig.or p value of less than 10%. That difference between the two conditions. That is, the variances is, the contact forces of the thumb tip and the index fingertip between condition means are likely due to chance and not are not influenced by the type of food to be consumed. likely due to the factor (independent variable) manipulation and vice versa for Sig. (2-tailed) lesser or equal to 0.05. From Table 7, for the bending motion of the index finger (BEND- 3.6. An Independent Samples t-Test of Bending Angles of Fingers with Respect to Different Eating Tools. An indepen- INDX), the Sig. (2-tailed) value is 0.034, which is less than dent samples t-test has been conducted using SPSS software 0.05. Thus, it can be concluded that there exists a statisti- cally significant difference in the means of bending motion to find whether the averaged bending angles of the thumb, index finger, and middle finger vary with respect to two dif- of the index finger, while using a fork (mean = 41.392) and a spoon (mean = 35.19) condition. That is, the bending ferent eating tool groups (a fork and a spoon). In this case, 10 Applied Bionics and Biomechanics Table 7: Independent samples t-test results for the bending motion data analysis. Levene’s test t-test for equality t-test for equality for equality of of means of means variances F Sig. t df Sig. (2-tailed) Equal variances assumed 13.38 0.000 2.20 348 0.029 BENDINDX Equal variances not assumed 2.13 265.58 0.034 Equal variances assumed 0.22 0.643 0.51 348 0.609 BENDMID Equal variances not assumed 0.52 307.26 0.606 Equal variances assumed 0.11 0.742 −2.02 348 0.045 BENDTHMB Equal variances not assumed −2.02 300.69 0.044 used (Sig. (2-tailed) = 0.494, reading from the bottom row). Table 8: Group statistics showing the Mean and SD for the contact Similar results have been obtained for the average contact force data analysis. force exerted by the thumb (FTHUMB); that is, the contact Cutlery Mean Std. deviation Std. error mean force exerted by the thumb is not influenced whether a fork Fork 0.90 0.46 0.04 or a spoon is used (Sig. (2-tailed) = 0.118, reading from the FINDX bottom row). Spoon 0.94 0.55 0.04 Fork 0.74 0.90 0.08 FTHMB 4. Discussion Spoon 0.60 0.77 0.05 Pearson correlation coefficient has been used to establish a relationship between the bending motion of the thumb, motion of the index finger is influenced by the type of cutlery index finger, and middle finger and the contact forces exerted used. Since results from (Table 6 showed that the mean of a by the thumb tip and the index fingertip during different eat- fork condition is higher than that of the spoon condition, it ing activities. The results revealed that for the cereal and soup can be concluded that the bending angles of the index finger eating activity using a spoon, the correlation coefficients using a fork are greater than its bending angles, while eating showed the same trend with the thumb motion, having the with a spoon. strongest positive correlation with the index fingertip force Similarly, for the bending motion of the middle finger and thumb-tip force, respectively. This can be attributed to (BENDMID), the results reveal that there exists no statisti- the fact that since, both the activities involve similar eating cally significant difference, whether eating with a spoon or action using a spoon, with only the food characteristics a fork; hence, the bending angles of the middle finger being different. Noodle and vegetable eating activities using remain unaffected, irrespective of the cutlery used (Sig. a fork showed different results. This can be due to the dif- (2-tailed = 0.609), reading from the top row). ferent eating action involved while picking up the food as The t-test results for the bending motion of the thumb the noodle involved rolling action of the fork. From the (BENDTHMB) (Tables 6 and 7) showed that there also exists correlation results, it can also be concluded that the middle a statistically significant difference in the means of bending finger motion has the weakest positive linear relationship motion of the thumb, while using a fork (mean = 33.02) and with the index fingertip and the thumb-tip force during a spoon (mean = 34.97) condition (Sig. (2-tailed = 0.045), all five eating activities, irrespective of the eating tools and reading from the top row). The results from (Table 6) showed food characteristics as compared to the thumb and index that the mean of a spoon condition is higher than that of the finger bending motion. fork condition; it can be concluded that the bending angles of A one-way ANOVA test has been done to compare the thumb using a spoon are greater than its bending angles bending motion of the thumb, index finger, and the middle while eating with a fork. finger and the contact forces exerted by the thumb and the 3.7. An Independent Samples t-Test of the Contact Forces index finger while eating different food types (cooked rice, Exerted by the Fingertips while Using Different Cutlery. A cereal with milk, vegetable salad, soup broth, and noodles). similar independent samples t-test (Tables 8 and 9) has been It can be concluded that the bending angles of thumb during performed to find if the average contact forces exerted by the the rice eating activity are relatively greater than cereal and thumb tip and the index fingertip vary with respect to differ- vegetable eating activities. Additionally, the bending motion ent eating tool groups (a fork and a spoon). Following the of the thumb during the vegetable eating activity is relatively same rules of interpreting the t-test as in the previous section, smaller than its bending motion during the noodle and soup it can be concluded from Tables 8 and 9 that for the average eating activity. Regarding the bending motion of the index contact force exerted by the index finger (FINDX), there finger, in all five eating activities, it can be concluded that exists statistically no significant difference in the means of during the noodle eating activity, the bending angles are the two conditions; that is, the contact force exerted by the relatively greater than cereal and soup eating activities, while index finger is not influenced whether a fork or a spoon is the bending angles of the index finger show no significant Applied Bionics and Biomechanics 11 Table 9: Independent samples t-test results for the contact force data analysis. Levene’s test t-test for equality of t-test for equality for equality of means of means variances F Sig. t df Sig. (2-tailed) Equal variances assumed 9.60 0.002 −0.66 348.00 0.509 FINDX Equal variances not assumed −0.68 330.37 0.494 Equal variances assumed 5.74 0.017 1.62 348.00 0.106 FTHMB Equal variances not assumed 1.57 265.59 0.118 statistical difference during the rice and vegetable eating Acknowledgments activities. The ANOVA results also revealed that the bending The authors would like to thank the Ministry of Education motion of the middle finger showed no significant statistical Malaysia for supporting this research under the Fundamental difference to different types of food. This means that the Research Grant Scheme (FRGS14-107-0348). bending motion of the middle finger is not varied much by food characteristics (solid or liquid or mixture of solid and liquid). The ANOVA results for the contact forces exerted References by the thumb and the index finger show that the forces are unaffected by the type of food. [1] S. K. Ostwald, M. P. Bernal, S. G. Cron, and K. M. Godwin, “Stress experienced by stroke survivors and spousal caregivers An independent samples t-test has been carried out to during the first year after discharge from inpatient rehabilita- compare the bending angles of the thumb, index finger, and tion,” Topics in Stroke Rehabilitation, vol. 16, no. 2, pp. 93– middle finger and the contact forces exerted by the thumb 104, 2009. and the index finger, respectively, while using different cut- [2] W.-K. Tang, C. G. Lau, V. Mok, G. S. Ungvari, and lery (fork and spoon). The results revealed that the bending K.-S. Wong, “Burden of Chinese stroke family caregivers: the angles of the index finger and the thumb are influenced by Hong Kong experience,” Archives of Physical Medicine and the type of cutlery unlike the middle finger which remains Rehabilitation, vol. 92, no. 9, article 21878218, pp. 1462– unaffected. Thus, we can conclude that the motion of the 1467, 2011. middle finger is not affected by the type of food characteris- [3] K. M. Godwin, S. K. Ostwald, S. G. Cron, and J. Wasserman, tics and the types of cutlery (fork/spoon) being used. The “Long-term health-related quality of life of stroke survivors independent samples t-test also revealed that the contact and their spousal caregivers,” Journal of Neuroscience Nursing, force exerted by the thumb and the index finger is not influ- vol. 45, no. 3, pp. 147–154, 2013. enced by the cutlery. Hence, it can be concluded that the con- [4] W. Duggleby, A. Williams, S. Ghosh et al., “Factors influencing tact force exerted by the index fingertip and the thumb tip are changes in health related quality of life of caregivers of persons not influenced by the different food characteristics nor by the with multiple chronic conditions,” Health and Quality of Life cutlery being used. Outcomes, vol. 14, no. 1, p. 81, 2016. [5] W. Mudzi, A. Stewart, and E. Musenge, “Caregiver strain and associated factors 12 months post stroke: impact of caregiver 5. Conclusion education,” Physiotherapy, vol. 101, no. 1, article e1055, 2015. [6] C. A. Gbiri, O. A. Olawale, and S. O. Isaac, “Stroke manage- In this study, human hand motion analysis was carried out ment: informal caregivers’ burdens and strians of caring for on five different eating activities with six subjects. The stroke survivors,” Annals of Physical and Rehabilitation Medi- ANOVA and t-test results revealed that the bending motion cine, vol. 58, no. 2, pp. 98–103, 2015. of the index finger and the thumb is affected with respect to [7] A. Westergren, I. R. Hallberg, and O. Ohlsson, “Nursing different food characteristics as well as the type of cutlery assessment of dysphagia among patients with stroke,” Scandi- used, that is, a fork and a spoon, whereas the bending motion navian Journal of Caring Sciences, vol. 13, no. 4, pp. 274–282, of the middle finger remains unaffected. In addition, the con- 1999. tact force exerted by the thumb tip and index fingertip [8] A. Westergren, M. Unosson, O. Ohlsson, B. Lorefält, and I. R. remains unaffected with respect to the different types of food Hallberg, “Eating difficulties, assisted eating and nutritional status in elderly (≥65 years) patients in hospital rehabilita- and cutlery used. These results can be useful in the future to tion,” International Journal of Nursing Studies, vol. 39, no. 3, differentiate hand motions dependent on different eating pp. 341–351, 2002. activities and the different cutlery (fork/spoon) used. It can [9] A. Westergren, “Detection of eating difficulties after stroke: a further be used in the development of a mathematical model systematic review,” International Nursing Review, vol. 53, of the hand for eating rehabilitation purposes. no. 2, pp. 143–149, 2006. [10] C. Jacobsson, K. Axelsson, P. O. Osterlind, and A. Norberg, “How people with stroke and healthy older people experience Conflicts of Interest the eating process,” Journal of Clinical Nursing, vol. 9, no. 2, The authors declare that they have no conflicts of interest. pp. 255–264, 2000. 12 Applied Bionics and Biomechanics [11] Z. Ju and H. Liu, “Human hand motion analysis with multi- sensory information,” IEEE/ASME Transactions on Mechatro- nics, vol. 19, no. 2, pp. 456–466, 2014. [12] R. A. R. C. Gopura, K. Kiguchi, and E. Horikawa, “A study on human upper-limb muscles activities during daily upper-limb motions,” International Journal of Bioelectromagnetism, vol. 12, no. 2, pp. 54–61, 2010. [13] X. Tang, Y. Liu, C. Lv, and D. Sun, “Hand motion classification using a multi-channel surface electromyography sensor,” Sen- sors, vol. 12, no. 12, pp. 1130–1147, 2012. [14] J. J. Cabibihan, I. Ahmed, and S. S. Ge, “Force and motion analyses of the human patting gesture for robotic social touch- ing,” in 2011 IEEE 5th International Conference on Cybernetics and Intelligent Systems (CIS), pp. 165–169, Qingdao, China, [15] J. Rosen, J. C. Perry, N. Manning, S. Burns, and B. Hannaford, “The human arm kinematics and dynamics during daily activ- ities - toward a 7 DOF upper limb powered exoskeleton,” ICAR '05 Proceedings, 12th International Conference on Advanced Robotics, 2005, 2005, pp. 532–539, Seattle, WA, USA, July, [16] J. Ah, E. Joo, P. Woo, H. Ram, J. Hyuk, and J. Nam, “Three- dimensional kinematic motion analysis of door handling task in people with mild and moderate stroke,” Physical Therapy Rehabilitation Science, vol. 5, pp. 143–148, 2016. [17] I. Aprile, M. Rabuffetti, L. Padua, E. Di Sipio, C. Simbolotti, and M. Ferrarin, “Kinematic analysis of the upper limb motor strategies in stroke patients as a tool towards advanced neuror- ehabilitation strategies: a preliminary study,” BioMed Research International, vol. 2014, Article ID 636123, 8 pages, 2014. [18] N. H. Adnan, K. Wan, A. B. Shahriman et al., “Measurement of the flexible bending force of the index and middle fingers for virtual interaction,” Procedia Engineering, vol. 41, pp. 388– 394, 2012. [19] N. Adnan, K. Wan, and A. Shahriman, “Accurate measure- ment of the force sensor for intermediate and proximal pha- langes of index finger,” International Journal of Computers Applications, vol. 45, no. 15, pp. 59–65, 2012. [20] A. G. Bluman, Elementary Statistics : A Step by Step Approach, McGraw-Hill, New York, NY, USA, 2012. [21] D. J. Biau, B. M. Jolles, and R. Porcher, “P value and the theory of hypothesis testing: an explanation for new researchers,” Clinical Orthopaedics and Related Research, vol. 468, no. 3, pp. 885–892, 2010. [22] R. A. Fisher, Statistical Methods for Research Workers, Springer, New York, NY, USA, 1992. International Journal of Advances in Rotating Machinery Multimedia Journal of The Scientific Journal of Engineering World Journal Sensors Hindawi Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 http://www www.hindawi.com .hindawi.com V Volume 2018 olume 2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Journal of Control Science and Engineering Advances in Civil Engineering Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com Journal of Journal of Electrical and Computer Robotics Engineering Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 VLSI Design Advances in OptoElectronics International Journal of Modelling & Aerospace International Journal of Simulation Navigation and in Engineering Engineering Observation Hindawi Hindawi Hindawi Hindawi Volume 2018 Volume 2018 Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com www.hindawi.com www.hindawi.com Volume 2018 International Journal of Active and Passive International Journal of Antennas and Advances in Chemical Engineering Propagation Electronic Components Shock and Vibration Acoustics and Vibration Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Applied Bionics and Biomechanics Hindawi Publishing Corporation

Human Hand Motion Analysis during Different Eating Activities

Loading next page...
 
/lp/hindawi-publishing-corporation/human-hand-motion-analysis-during-different-eating-activities-etongAaOrx
Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2018 Zakia Hussain et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1176-2322
eISSN
1754-2103
DOI
10.1155/2018/8567648
Publisher site
See Article on Publisher Site

Abstract

Hindawi Applied Bionics and Biomechanics Volume 2018, Article ID 8567648, 12 pages https://doi.org/10.1155/2018/8567648 Research Article Zakia Hussain , Norsinnira Zainul Azlan , and Arif Zuhairi bin Yusof Department of Mechatronics Engineering, International Islamic University Malaysia, Gombak, 53100 Kuala Lumpur, Malaysia Correspondence should be addressed to Norsinnira Zainul Azlan; sinnira@iium.edu.my Received 6 October 2017; Revised 24 November 2017; Accepted 20 December 2017; Published 4 February 2018 Academic Editor: Laurence Cheze Copyright © 2018 Zakia Hussain et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The focus of this research is to analyse both human hand motion and force, during eating, with respect to differing food characteristics and cutlery (including a fork and a spoon). A glove consisting of bend and force sensors has been used to capture the motion and contact force exerted by fingers during different eating activities. The Pearson correlation coefficient has been used to show that a significant linear relationship exists between the bending motion of the fingers and the forces exerted during eating. Analysis of variance (ANOVA) and independent samples t-tests are performed to establish whether the motion and force exerted by the fingers while eating is influenced by the different food characteristics and cutlery. The middle finger motion showed the least positive correlation with index fingertip and thumb-tip force, irrespective of the food characteristics and cutlery used. The ANOVA and t-test results revealed that bending motion of the index finger and thumb varies with respect to differing food characteristics and the type of cutlery used (fork/spoon), whereas the bending motion of the middle finger remains unaffected. Additionally, the contact forces exerted by the thumb tip and index fingertip remain unaffected with respect to differing food types and cutlery used. 1. Introduction during eating, an in-depth knowledge of hand motion during eating is vital. Upper limb disability is one of the major adversities faced by Hand motion during eating is highly dexterous and is poststroke patients. The resulting loss of mobility in these subject to the type of food ingested and the type of cutlery patients reduces their ability to perform normal activities used. Analysing hand motion can be complicated due to its of daily living (ADL), preventing them from leading a nor- highly articulate nature. A human hand consists of 27 bones mal life and hence reducing their quality of life. These and 35 muscles, of which 17 are intrinsic muscles (located in patients are highly dependent on their caregivers (usually a the palm) and 18 are extrinsic muscles (located in the fore- arm). With roughly 30 degrees of freedom (DOFs), this com- spouse or friend) who perform most of their basic ADL, such as eating, bathing, and grooming, which gradually has plex structure can perform intricate tasks, which require a negative impact on the mental and physical state of the dexterity. During the past few years, hand motion analysis caregiver [1–6]. has gained the attention of the researchers working in the Eating is one of the fundamental activities of survival for field of rehabilitation, human-computer interaction (HCI), all living beings. Dysphagia and other eating difficulties are and robotics. also common among poststroke patients which can lead to Hand motion analysis enables researchers to gather data complications, such as malnutrition, dehydration, suffoca- such as the force applied by the fingers, different joint angles tion, and eventually death [7–10]. Over the past decade, of the hand, and velocity, while performing different activi- numerous robotic rehabilitation systems have been devel- ties. Analysing the motion and force of the hand during var- oped to assist impaired patients regain their hand functions. ious eating activities can help in formulating a model, which Such robotic systems must have the capability to replicate in turn can be useful in developing a rehabilitation robot for human hand function during any ADL. To develop a rehabil- assisted eating. Several studies have been conducted to ana- itation system meant specifically to regain the hand function lyse the motion of the hand and upper limb while performing 2 Applied Bionics and Biomechanics discussion of the data analysis results in the previous section, different daily activities of living. Ju and Liu [11], Gopura et al. [12], and Tang et al. [13] have successfully analysed and lastly, the conclusion is drawn in Section 5. and classified different human hand motions while perform- ing basic daily activities, such as hair combing and recogniz- 2. Experimental Method ing multiple hand gestures, using electromyography (EMG). In EMG analysis, tiny electrodes, when placed on human 2.1. Experimental Setup. A prototype glove has been used skin, detect and record the electrical signals transmitted by to analyse the motion of hand during eating (Figure 1). the motor neurons responsible for activating muscle contrac- The glove for hand has been designed as an instrument tion. Ju and Liu [11] used a framework of multiple sensor to measure the angle of the index finger, middle finger, and integration of CyberGlove, Finger TPS pressure sensors, thumb. The glove is developed with three flexible bend sen- and Trigno wireless EMG sensors to capture hand gestures, sors (Spectra Symbol, 4.5 inches) for measuring the angles contact forces, and muscle contraction signals from various of the index finger, middle finger, and thumb (Figure 2). hand motions, while performing 10 basic grasping activities, These bend sensors act as variable resistors which, when such as holding and lifting a dumbbell and opening and clos- flexed, increase the resistance across the sensor. Force sensors ing a pen box, using five fingers. (FlexiForce™, A201) are attached to the finger tip of the index Cabibihan et al. [14] explored the human patting ges- and thumb to measure the force exerted by the thumb and ture for analysing the amount of force applied to regions index finger, during eating process, since only the index fin- of the hand and the angular motion of finger joints so as ger and thumb are involved in holding the spoon/fork during to incorporate them into a humanoid robot, in order to any eating activity. imitate this gesture. Similarly, the kinematics and dynamics The data from the glove is recorded using MATLAB 2015 of the human arm, during 24 daily activities (such as eating through serial communication with Arduino. (Figure 3) using a spoon and a fork, drinking with a cup, and washing demonstrates the hardware setup of the bend sensors and the face) were studied by Rosen et al. [15] to develop a 7- the force sensors. DOF powered exoskeleton for the upper limb. Ah et al. [16] performed human hand motion analysis while turning 2.2. Data Acquisition. Six healthy, right-handed subjects a door knob. including three males and three females, age ranging from Aprile et al. [17] dedicated an entire study to analyse the 24 to 30 years and an average weight of 65 kgs, volunteered upper limb motion in stroke patients while performing a for this study. Five eating activities were performed, to ana- drinking task, which included reaching for the glass, bringing lyse the hand motion, while using different eating cutlery it to the mouth, and putting it back on the table. Adnan et al. (spoon and fork) and food types (including solids and liq- [18] developed a low-cost DataGlove using a flexible bend uids). The type of food involved in the eating activities sensor to recognize various human finger activities. In addi- included cooked rice, milk cereal, salad with chunks of vege- tion, the analytical mathematical model and analysis of vari- tables, noodles, and a clear soup broth. A plastic spoon and a ance (ANOVA) was established to predict the force induced steel fork were used during the activity. Each activity was per- at the flexible force sensor by the human finger using the low- formed three times by each participant, with each trial lasting cost DataGlove [19]. seven seconds and while sitting on a chair with food on the Some previous work on hand analysis is summarized in table. The five activities performed were as follows: Table 1. Despite many studies on hand motion (Table 1), to (1) Eating rice (solid) with a spoon the best of our knowledge, there has not been a study dedi- cated to the analysis of hand motion while eating different (2) Eating soup broth (liquid) with a spoon types of food and using different cutlery. It is important to consider the food characteristics and the amount of force (3) Eating cereal with milk (mixture of solid and liquid) exerted by the hand during eating to enhance the develop- using a spoon ment of robotic rehabilitation systems for this activity. (4) Eating vegetable salad (solid) using a fork Therefore, this paper presents human hand motion anal- ysis, focusing on the thumb, index finger, and middle finger (5) Eating noodles (solid) using a fork during eating. The motion of these three fingers and the force they exert during eating is studied with respect to the type of Subjects were trained before performing the activities on food (liquid, solid) and the cutlery used. An experiment has how to grasp the cutlery and eat using it, while wearing the been conducted involving five different food types and using glove. For eating noodles, subjects were asked to roll over two different types of cutlery (fork and spoon) to study their the noodles on the fork and then eat. Each trial of the eating effect on hand motion. ANOVA and t-test analysis has been activities performed consisted of four main events (Figure 4). conducted to study the influence of these factors on the finger The first event in each eating activity was the origin or start- motion and force during eating. The paper is organized as ing point, which occurred when the subject kept the glove follows: Section 2 presents the method of experimentation rested horizontally on the table and bend sensors with almost employed which is subdivided into two subsections: Experi- no bend. The second event called event A occurred, when the mental Setup and Data Acquisition. Section 3 presents the subject holding a spoon or fork digs in to the food and brings data analysis and results obtained while eating different types it towards the mouth to eat. The third event known as event B of food and using different cutlery. Section 4 presents the occurred, when the subject during eating maintains the grip Applied Bionics and Biomechanics 3 Table 1: Highlights of the previous contributions to human motion analysis. Data Number Authors Objective Focus of study acquisition Results/findings Activity method Strong correlations between To correlate the muscle Ten in-hand Human hand motion muscle signals, contact forces, signals with contact forces EMG sensor, manipulations Ju and Liu analysis with and finger trajectories. 1 and finger trajectories & force sensor & like holding & [11] multisensory Fuzzy Gaussian mixture DataGlove lifting a motion recognition using information models (FGMMs) used for muscle signals dumbbell motion recognition To analyse upper-limb EMG Relationships between the Basic motions muscle activities during Human upper-limb electrodes, upper limb motions & and the Gopura basic upper-limb motion, muscle activities 2 VICON activity levels of main selected daily et al. [12] to design power-assist during daily upper- motion muscles have been activities of robotic exoskeleton limb motions capture system established upper-limb systems Hand motion Experimental results showed To classify multiple hand Tang et al. classification using a that the success rate for the 11 hand 3 gestures using three sEMG sensors [13] multichannel surface identification of all the 11 gestures different methods sEMG sensor gestures is significantly high To analyse the gesture, the amount of force applied Human patting gesture CyberGlove II The sensitive regions on the Cabibihan Human 4 on regions of the hand, analysis for robotic FingerTPS hand while performing pat et al. [14] patting gesture and the angular motion of social touching sensors have been identified finger joints VICON The results indicated that the To study the kinematics The human arm motion various joints’ kinematics and Rosen and the dynamics of the kinematics and 5 capture system dynamics change 24 ADL human arm during daily dynamics during daily et al. [15] &reflective significantly based on the activities activities markers nature of the task To evaluate motor control 3D kinematic motion VICON abilities between the analysis of door motion Comparisons have been Ah et al. Door handling 6 groups of people with mild handling task in capture system drawn between healthy, mild [16] task and moderate arm people with mild and &reflective & moderate stroke patients impairments moderate stroke markers To analyse, using motion Comparisons have been Smart motion analysis, the qualitative Kinematic analysis of drawn between stroke & Aprile capture 7 and quantitative upper the upper limb motor healthy control group while Drinking task et al. [17] optoelectronic limb motor strategies in strategies in stroke reaching out for the glass to system stroke patients drink Low-cost Measurement of the Sign language To develop a low-cost DataGlove by flexible bending force The DataGlove developed translation Adnan DataGlove, able to using the 8 of the index and can measure several human (letters A, B, C, et al. [18] recognize the different flexible middle fingers for degrees of freedom (DoFs) D, F & K and finger activities bending virtual interaction number 8) sensor Accurate An analytical mathematical measurement of force model and ANOVA has been To find the correlations Flexiforce Any finger Adnan by the force sensor for established to predict the 9 between the forces of pressure gripping et al. [19] intermediate and force induced at the flexible finger phalanges sensors activity proximal phalanges of force sensor and the human index finger finger of low-cost DataGlove on the cutlery. The final event is known as event C, when the 3. Data Analysis and Results subject releases the cutlery after eating and goes back again to the origin, that is, the subject after finishing the eating 3.1. Bending Finger Motion Trajectories. The motion trajecto- brought his/her hand back to rest on the table. Throughout ries captured by the bend sensors for the thumb, index finger, the experiment, subjects were asked to keep their elbows and middle finger during five different eating activities are rested on the table. shown (Figure 5) with the four important events identified 4 Applied Bionics and Biomechanics Force sensors Bending sensors (a) (b) Figure 1: (a) The location of bend sensors on the thumb, middle finger, and index finger. (b) The location of force sensors on the thumb and the index finger. (a) (b) (c) Figure 2: (a) The index finger angle, (b) the middle finger angle, and (c) the thumb angle, measured by the bend sensors. Data from glove (bend sensors & Arduino MATLAB Computer force sensors) Figure 3: Hardware setup of the bend and force sensors for hand motion analysis. Origin Event A (a) (b) Event C Event B (c) (d) Figure 4: Four main events identified during each eating activity: (a) origin, (b) event A, (c) event B, (d) Event C. Applied Bionics and Biomechanics 5 umb motion trajectory Index finger motion trajectory Event A Event A 70 Event C Event C Origin Event B Origin Event B Origin Origin 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 Time (s) Time (s) Rice Veg Rice Veg Cereal Soup Cereal Soup Noodle Noodle (a) (b) Middle finger motion trajectory Event A Event C Event B Origin Origin 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 Time (s) Rice Veg Cereal Soup Noodle (c) Figure 5: (a) Thumb motion trajectories, (b) index finger motion trajectories, and (c) middle finger motion trajectories obtained from the bend sensor for five different eating activities. on the trajectory. During all five eating activities (rice, cereal, maximum averaged angle of 121.8 ; for noodle eating activ- and soup with a spoon, noodle and vegetable using a fork), ity, the ROM is from a minimum averaged angle 17 to a the averaged range of motion (ROM) for the thumb ranged maximum averaged angle of 115 ; for cereal with milk ° ° from a minimum of 19.5 activity using a spoon, the ROM is from a minimum aver- to a maximum of 59.1 (referring ° ° ° to Figure 2). The origin is around 19 , although the subjects aged angle of 17 to a maximum averaged angle of 111 ; kept their hands horizontally at rest on the table; this can for eating vegetables using a fork, the ROM is from a mini- ° ° be due to the sensor fatigue while doing the activities repeat- mum averaged angle of 17 to a maximum of 103 ; and for edly (Figure 5(a)). the soup broth eating activity using a spoon, the ROM is The ROM of the index finger during eating rice with a from a minimum of 17 to a maximum averaged angle of ° ° spoon is from a minimum averaged angle of 7 to a maxi- 120.6 (Figure 5(c)). mum averaged angle of 90 ; for the noodle eating activity, From the graphs (Figures 5(a)–5(c)), it can be observed the ROM is from a minimum averaged angle of 7 to a max- that as event A starts (at around 0.5 seconds), the bending imum averaged angle of 93 ; for cereal with milk activity angles of the fingers start increasing to grip the cutlery and reach a maximum value, while bringing the food to the using a spoon, the ROM is from a minimum averaged angle ° ° of 7 to a maximum averaged angle of 75 ; for vegetable eat- mouth. During event B (lasting around 3 seconds), the ing activity using a fork, the ROM is from a minimum aver- magnitude of the angles remains steady, while the subject ° ° aged angle of 9 to a maximum averaged angle of 83 ; and for maintains the grip on the cutlery during eating. The magni- the soup broth eating activity using a spoon, the ROM is from tude of the bending angles starts decreasing during event C ° ° a minimum of 7 to maximum averaged angle of 72 . The ori- (lasting around 2 seconds), when the subject releases the gin in all activities is around 7 (Figure 5(b)). grip off the cutlery by putting it back into the dish and pro- The ROM of the middle finger whist eating rice with ceeding towards the origin, when the bending angles again ° ° a spoon is from a minimum averaged angle of 18 to a settle at around 20 . Average angles (º) Average angles (º) Average angles (º) 6 Applied Bionics and Biomechanics Force exerted by index fingertip Force exerted by thumb tip 3.00E + 00 2.50E + 00 2.00E + 00 3 1.50E + 00 1.00E + 00 5.00E − 01 0.00E + 00 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Time (s) Time (s) Trial 1 Trial 1 Trial 2 Trial 2 Trial 3 Trial 3 (a) (b) Figure 6: (a) Three trials of thumb-tip force and (b) index fingertip force captured by the force sensor during vegetable eating activity. Force exerted by thumb tip Force exerted by index fingertip Event C 2.5 Event A Event A Event C 2.5 1.5 1.5 Origin Origin Event B Event B Origin 0.5 0.5 Origin 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 Time (s) Time (s) Rice Veg Rice Veg Cereal Soup Cereal Soup Noodle Noodle (a) (b) Figure 7: (a) Thumb-tip force and (b) index fingertip force recorded by the force sensor during five different eating activities. 3.2. Contact Force Trajectories. The force sensors attached subject is putting the spoon/fork back into the dish, eventu- to the prototype glove measured the force exerted by the ally, coming back to the origin, when the subject rests his/ thumb tip and the index fingertip, while performing five her hand on table again. During the noodle eating activity different eating activities, since only the thumb and index using a fork, the thumb tip exerts a maximum average force finger are involved in holding the spoon or fork while eating. of 2.7 N, which is the highest of all other eating activities. To check the repeatability of the force sensors used, a subject Since during the noodle eating activity, the subjects were performing three trials of the vegetable eating activity is asked to roll over the noodles on the fork; hence, the force shown in Figure 6. trajectory shows some minor fluctuations and is longer than From Figure 6, it can be observed that the force sensor other eating activities. For rice eating activity using a spoon, a measurements, attached to the index fingertip and the thumb maximum average force of 2.4 N; for cereal with milk eating tip, demonstrate quite consistent results. activity using a spoon, a maximum average force of 2.2 N; (Figures 7(a) and 7(b)) demonstrate the force exerted by for vegetable eating activity using a fork, a maximum average the thumb tip and the index fingertip with the four main force of 2.4 N; and for soup broth eating activity using a events identified on the graphs. During the origin, the hand spoon, a maximum average force of 2.0 N is exerted by the is lying horizontally on the table, and as such, no force is thumb tip (Figure 7(a)). During the cereal with milk eating activity using a spoon, exerted by the thumb tip/index fingertip, but as event A starts and the subject grips a spoon or fork between the thumb and the index fingertip exerts a maximum average force of 2.4 N, index finger, the magnitude of force increases and during which is the highest of all other eating activities. For rice eat- event C and a maximum force is reached, when the subject ing activity using a spoon, a maximum average force of 2.0 N; is digging into the food or trying to get the food on the spoon for noodle eating activity using a fork, maximum average force of 2.0 N; for vegetable eating activity using a fork, a or fork. The force then starts to decrease in event C, as the Force (n) Force (n) Force (n) Force (n) Applied Bionics and Biomechanics 7 Table 2: Averaged Pearson coefficient of bending finger angles and force exerted by fingers. Rice (spoon) Cereal & Milk (spoon) Soup (spoon) Vegetable (fork) Noodle (fork) Index Middle Thumb Index Middle Thumb Index Middle Thumb Index Middle Thumb Index Middle Thumb FINDEX 0.93 0.87 0.78 0.89 0.84 0.95 0.90 0.87 0.94 0.89 0.81 0.74 0.85 0.73 0.89 FTHUMB 0.90 0.84 0.81 0.89 0.83 0.94 0.90 0.89 0.94 0.90 0.83 0.80 0.86 0.72 0.88 Table 3: Analysis of variance summary table. maximum average force of 2.2 N; and for soup broth eating activity using a spoon, a maximum average of 2.1 N force is Sum of Mean Sig./p df exerted by the index fingertip (Figure 7(b)). squares square value Between 3.3. Correlations between Bending Angles and the Contact 893.33 4 223.33 0.023 groups Forces of Fingers. The bending finger angle data and the con- BENDTHMB Within tact force data captured by the sensors has been used to find 26855.55 345 77.84 groups the correlations between the bending angles of the thumb, Total 27748.88 349 index finger, and middle finger and the contact forces exerted by the thumb and index finger, during eating activities using Between 5765.47 4 1441.37 0.074 the Pearson product moment correlation coefficient (PPMC). groups Pearson correlation coefficient (r) measures the strength BENDINDX Within 230984.22 345 669.52 and direction of a linear relationship between two variables. groups The SPSS statistics software package has been used to per- Total 236749.69 349 form this analysis. It takes values ranging from +1 to −1; Between 3236.77 4 809.19 0.609 r =+1 implies a strong positive linear relationship between groups the variables, while r = 1 implies a strong negative linear rela- BENDMID Within tionship, and r =0 implies no linear relationship between the 413100.47 345 1197.39 groups variables. Equation (1) gives the formula for computing the Total 416337.24 349 Pearson correlation coefficient [20]. n〠 xy − 〠x 〠y groups. SPSS Statistics software package has been used to r = , 1 perform ANOVA in this study. In this study, ANOVA has 2 2 2 2 n 〠x − 〠x n 〠y − 〠y been used to determine whether the bending motion of the index finger (BENDINDX), middle finger (BENDMID), and thumb (BENDTHMB)differed based on different groups of where n = number of data pairs. food types (cereal, rice, vegetable, noodle, and soup) or not The averaged Pearson coefficients of different bending (Tables 3 and 4). finger angles and the forces exerted by the index finger and The df column in Table 3 means degrees of freedom, thumb, for all six subjects involved in the experiment, are which is the division of Sum of squares by Mean square values shown in (Table 2). in the ANOVA summary table. The Sum of squares is the The most significant coefficients are highlighted sum of Between groups and Within groups. The Sig. column (Table 2). From the results (Table 2), both the soup (liquid) denotes the p value which represents the probability of find- and cereal (solid and liquid) eating activities which are per- ing an effect equal to or greater than the one observed consid- formed using a spoon have similar results, where both index ering the null hypothesis to be true. The null hypothesis here fingertip force and the thumb-tip force have shown the stron- signifies that there is no significant difference in the bending gest positive linear relationship with the average bending angles of the thumb, index finger, and middle finger with motion of the thumb. Noodle eating activity has shown sim- respect to different types of food groups. ilar results with the soup and cereal activities. Moreover, dur- The lower the p value, the more likely the null hypothesis ing the vegetable and rice eating activities, the force exerted is rejected (preferably less than 0.05, while 0.10 is also by the thumb tip and the index fingertip have the strongest accepted but as a weak evidence). The p value thus provides positive correlation with the averaged index finger motion. a quantitative strength of evidence against the null hypothe- During all the eating activities, the middle finger motion sis stated [21, 22]. From (Table 3), it can be concluded that has the weakest linear relationship with the index fingertip for the average bending motion of the thumb and the index and thumb-tip force as compared to the bending motion of finger across the five groups of food (soup, rice, noodle, the index finger and thumb. cereal, and vegetable salad), there is a statistically significant 3.4. One-Way ANOVA of the Bending Angles of Fingers with difference at 5% and 10%, respectively, whereas for the aver- Respect to Different Types of Food. The one-way analysis of aged bending motion of the middle finger, there is no statis- tically significant difference, while eating different types of variance or ANOVA is a statistical comparison test used to determine whether there are any statistically significant dif- food (p > 10%). In simple words, it can be concluded from the ANOVA results that the bending motion of the thumb ferences between the means of two or more independent 8 Applied Bionics and Biomechanics Table 4: A least significant difference post hoc test using SPSS An LSD (least significant difference) post hoc test has software package. been carried out to distinguish eating precisely which type of food group (noodle, soup, rice, vegetables, and cereal) Dependent (I) (J) Mean difference Sig. the variances occur in the bending motions of the thumb variable foodtype foodtype (I − J) and the index finger (Table 4). To check for the vari- BENDTHMB ances, Mean difference (I-J) and Sig. (significance/p value) Rice −2.66 0.075 column of Table 4 is considered. The LSD results for the bending angles of the thumb from Table 4 can be sum- Veg 2.33 0.119 1 Cereal marised as follows: Noodle −0.60 0.687 Soup −0.61 0.684 (1) The bending angles of thumb during cereal eating Cereal 2.66 0.075 activity are smaller than its bending angles during 4.99 Veg 0.001 rice eating activity at a statistically significant differ- 2 Rice Noodle 2.06 0.168 ence of 10% (p =0 075, read row 1). Soup 2.06 0.169 (2) On the contrary, the bending angles of thumb during Cereal −2.33 0.119 vegetable, noodle, and soup eating activities show sta- −4.99 Rice 0.001 tistically no significant difference with the bending 3 Veg Noodle −2.93 0.050 motion of thumb during cereal eating activity. In Soup −2.94 0.050 other words, the bending motion of the thumb dur- ing vegetable, noodle, and soup eating activities does Cereal 0.60 0.687 not show any variance with respect to the cereal eat- Rice −2.06 0.168 4 Noodle ing activity. Veg 2.93 0.050 (3) The bending angles of thumb during rice eating Soup 0.00 0.998 activity are again greater than its bending angles dur- Cereal 0.61 0.684 ing the vegetable eating activity, at a statistically sig- Rice −2.06 0.169 5 Soup nificant difference of 1%, while the bending angles 2.94 Veg 0.050 of thumb during noodle and soup activities show sta- Noodle 0.00 0.998 tistically no significant difference with the bending BENDINDX motion of thumb during rice eating activity. Rice −3.29 0.452 (4) Similarly, the bending angles of thumb during Veg −3.33 0.447 vegetable eating activity are lesser than its bending 6 Cereal −9.97 Noodle 0.023 angles during noodle and soup eating activities at a statistically significant difference of 10% and 5%, Soup 1.96 0.655 respectively. Cereal 3.29 0.452 Veg −0.04 0.993 (5) Finally, the bending motion of the thumb during 7 Rice Noodle −6.68 0.127 noodle eating activity has statistically no significant difference with the bending motion of the thumb Soup 5.25 0.231 during soup eating activity. Cereal 3.33 0.447 Rice 0.04 0.993 Therefore, from the LSD post hoc results, it can be 8 Veg Noodle −6.65 0.129 concluded that the bending motion of the thumb during Soup 5.28 0.228 rice and vegetable eating activity is maximum as compared to other eating activities and the bending motion of the Cereal 9.97 0.023 thumb does not show much variance during noodle and Rice 6.68 0.127 9 Noodle soup eating activities. Veg 6.65 0.129 Similarly, the ANOVA results for bending angles of index 11.93 Soup 0.007 finger (Table 4) can be summarised as follows: Cereal −1.96 0.655 Rice −5.25 0.231 (1) The bending angles of index finger during cereal eat- 10 Soup Veg −5.28 0.228 ing activity are smaller than its bending angles during −11.93 the noodle eating activity, at a statistically significant Noodle 0.007 difference of 5%. The mean difference is significant at the 0.05 level. (2) Contrary, the bending angles of index finger during and the index finger is influenced by the type of food, rice, vegetable, and soup eating activity show statis- whereas the bending motion of the middle finger is not tically no significant difference with the bending motion of index finger during cereal eating activity. affected by type of food. Applied Bionics and Biomechanics 9 Table 5: ANOVA summary table. Table 6: Group statistics showing the mean and standard deviation (SD) for the bending motion data analysis. Sum of Mean Sig./p df squares square value Cutlery Mean Std. deviation Between Fork 41.39 28.29 0.30 4 0.08 0.892 BENDINDX groups Spoon 35.19 24.19 FINDX Within Fork 58.12 33.62 94.05 345 0.27 groups BENDMID Spoon 56.19 35.20 Total 94.35 349 Fork 33.02 8.81 Between BENDTHMB 3.54 4 0.88 0.273 Spoon 34.97 8.92 groups FTHMB Within 236.09 345 0.68 groups t-test is conducted instead of ANOVA because here, the fac- Total 239.63 349 tor (independent variable) which is the cutlery has only two groups (fork/spoon), but for conducting ANOVA, it must be more than two; hence, an independent samples t-test has (3) The bending angles of the index finger during rice eating activity show statistically no significant differ- been conducted. The results of the independent samples t-test are shown ence with the bending motion of the index finger dur- ing the vegetable, noodle, and soup eating activities. in Tables 6 and 7. To interpret the results from the t-test table (Table 7), the large column labelled Levene’s test for equality (4) During the vegetable eating activity, the bending of variances is checked first. This is a test that determines if motion of the index finger shows statistically no sig- the two conditions (a fork and a spoon) have about the same nificant difference with the bending motion of the or different amounts of variability between scores. Under this index finger during noodle and soup eating activities. column, the Sig. p value column is considered. This Sig. value determines which row to consider, either the Equal variances (5) Finally, during the noodle eating activity, the bending assumed or the Equal variances not assumed row. If the Sig. angles of the index finger are greater than its bending value is greater than 0.05, read from the top row, which angles, during soup eating activity at a statistically means that the variability in the two conditions is about the significant difference of 10%. same. That is, the scores in one condition (fork) do not vary Thus, from the LSD post hoc test results, it can be sum- much more than the scores in the second condition (spoon). marised that the bending motion of the index finger during Put scientifically, it means that the variability in the two con- noodle eating activity is highest as compared to the other four ditions is not significantly different and vice versa, if the Sig. eating activities. Additionally, the bending motion of the value is lesser or equal to 0.05. In the latter case, read from the index finger does not show any significant variance with bottom row, that the variability in the two conditions is not respect to soup, vegetable, and rice eating activities. the same. That is, the scores in one condition vary much more than the scores in the second condition. Scientifically, 3.5. One-Way ANOVA of the Forces Exerted by the Finger it means that the variability in the two conditions is signifi- Tips with Respect to Different Types of Food. The one-way cantly different. ANOVA technique has been again used to determine if there From Table 7, for the bending motion of the index finger exists a statistically significant difference in the averaged (BENDINDX), the p value is less than 0.05 (0.000), that is, forces exerted by the thumb tip (FTHMB) and the index fin- reading from the bottom row, which reveals that the variabil- gertip (FINDX), based on the different groups of food type ity in the two conditions (fork and spoon) is not the same. (Table 5). From the ANOVA results, it can be concluded that After finding the row to read (bottom row), now the results for both forces exerted by the thumb and the index finger, of t-test can be found in the column labelled t-test for equality there exists statistically no significant difference during vari- of mean by considering the Sig. (2-tailed) column under it. ous eating activities (p > 10%). As seen from the Sig. column This Sig. (2-tailed) column determines if the two conditions’ in Table 5, for both index finger and the thumb, the Sig. is means are statistically different. If the Sig. (2-tailed) is greater 0.892 (89.2%) and 0.273 (27.3%), respectively, which is far than 0.05, this means that there is no statistically significant greater than the desired Sig.or p value of less than 10%. That difference between the two conditions. That is, the variances is, the contact forces of the thumb tip and the index fingertip between condition means are likely due to chance and not are not influenced by the type of food to be consumed. likely due to the factor (independent variable) manipulation and vice versa for Sig. (2-tailed) lesser or equal to 0.05. From Table 7, for the bending motion of the index finger (BEND- 3.6. An Independent Samples t-Test of Bending Angles of Fingers with Respect to Different Eating Tools. An indepen- INDX), the Sig. (2-tailed) value is 0.034, which is less than dent samples t-test has been conducted using SPSS software 0.05. Thus, it can be concluded that there exists a statisti- cally significant difference in the means of bending motion to find whether the averaged bending angles of the thumb, index finger, and middle finger vary with respect to two dif- of the index finger, while using a fork (mean = 41.392) and a spoon (mean = 35.19) condition. That is, the bending ferent eating tool groups (a fork and a spoon). In this case, 10 Applied Bionics and Biomechanics Table 7: Independent samples t-test results for the bending motion data analysis. Levene’s test t-test for equality t-test for equality for equality of of means of means variances F Sig. t df Sig. (2-tailed) Equal variances assumed 13.38 0.000 2.20 348 0.029 BENDINDX Equal variances not assumed 2.13 265.58 0.034 Equal variances assumed 0.22 0.643 0.51 348 0.609 BENDMID Equal variances not assumed 0.52 307.26 0.606 Equal variances assumed 0.11 0.742 −2.02 348 0.045 BENDTHMB Equal variances not assumed −2.02 300.69 0.044 used (Sig. (2-tailed) = 0.494, reading from the bottom row). Table 8: Group statistics showing the Mean and SD for the contact Similar results have been obtained for the average contact force data analysis. force exerted by the thumb (FTHUMB); that is, the contact Cutlery Mean Std. deviation Std. error mean force exerted by the thumb is not influenced whether a fork Fork 0.90 0.46 0.04 or a spoon is used (Sig. (2-tailed) = 0.118, reading from the FINDX bottom row). Spoon 0.94 0.55 0.04 Fork 0.74 0.90 0.08 FTHMB 4. Discussion Spoon 0.60 0.77 0.05 Pearson correlation coefficient has been used to establish a relationship between the bending motion of the thumb, motion of the index finger is influenced by the type of cutlery index finger, and middle finger and the contact forces exerted used. Since results from (Table 6 showed that the mean of a by the thumb tip and the index fingertip during different eat- fork condition is higher than that of the spoon condition, it ing activities. The results revealed that for the cereal and soup can be concluded that the bending angles of the index finger eating activity using a spoon, the correlation coefficients using a fork are greater than its bending angles, while eating showed the same trend with the thumb motion, having the with a spoon. strongest positive correlation with the index fingertip force Similarly, for the bending motion of the middle finger and thumb-tip force, respectively. This can be attributed to (BENDMID), the results reveal that there exists no statisti- the fact that since, both the activities involve similar eating cally significant difference, whether eating with a spoon or action using a spoon, with only the food characteristics a fork; hence, the bending angles of the middle finger being different. Noodle and vegetable eating activities using remain unaffected, irrespective of the cutlery used (Sig. a fork showed different results. This can be due to the dif- (2-tailed = 0.609), reading from the top row). ferent eating action involved while picking up the food as The t-test results for the bending motion of the thumb the noodle involved rolling action of the fork. From the (BENDTHMB) (Tables 6 and 7) showed that there also exists correlation results, it can also be concluded that the middle a statistically significant difference in the means of bending finger motion has the weakest positive linear relationship motion of the thumb, while using a fork (mean = 33.02) and with the index fingertip and the thumb-tip force during a spoon (mean = 34.97) condition (Sig. (2-tailed = 0.045), all five eating activities, irrespective of the eating tools and reading from the top row). The results from (Table 6) showed food characteristics as compared to the thumb and index that the mean of a spoon condition is higher than that of the finger bending motion. fork condition; it can be concluded that the bending angles of A one-way ANOVA test has been done to compare the thumb using a spoon are greater than its bending angles bending motion of the thumb, index finger, and the middle while eating with a fork. finger and the contact forces exerted by the thumb and the 3.7. An Independent Samples t-Test of the Contact Forces index finger while eating different food types (cooked rice, Exerted by the Fingertips while Using Different Cutlery. A cereal with milk, vegetable salad, soup broth, and noodles). similar independent samples t-test (Tables 8 and 9) has been It can be concluded that the bending angles of thumb during performed to find if the average contact forces exerted by the the rice eating activity are relatively greater than cereal and thumb tip and the index fingertip vary with respect to differ- vegetable eating activities. Additionally, the bending motion ent eating tool groups (a fork and a spoon). Following the of the thumb during the vegetable eating activity is relatively same rules of interpreting the t-test as in the previous section, smaller than its bending motion during the noodle and soup it can be concluded from Tables 8 and 9 that for the average eating activity. Regarding the bending motion of the index contact force exerted by the index finger (FINDX), there finger, in all five eating activities, it can be concluded that exists statistically no significant difference in the means of during the noodle eating activity, the bending angles are the two conditions; that is, the contact force exerted by the relatively greater than cereal and soup eating activities, while index finger is not influenced whether a fork or a spoon is the bending angles of the index finger show no significant Applied Bionics and Biomechanics 11 Table 9: Independent samples t-test results for the contact force data analysis. Levene’s test t-test for equality of t-test for equality for equality of means of means variances F Sig. t df Sig. (2-tailed) Equal variances assumed 9.60 0.002 −0.66 348.00 0.509 FINDX Equal variances not assumed −0.68 330.37 0.494 Equal variances assumed 5.74 0.017 1.62 348.00 0.106 FTHMB Equal variances not assumed 1.57 265.59 0.118 statistical difference during the rice and vegetable eating Acknowledgments activities. The ANOVA results also revealed that the bending The authors would like to thank the Ministry of Education motion of the middle finger showed no significant statistical Malaysia for supporting this research under the Fundamental difference to different types of food. This means that the Research Grant Scheme (FRGS14-107-0348). bending motion of the middle finger is not varied much by food characteristics (solid or liquid or mixture of solid and liquid). The ANOVA results for the contact forces exerted References by the thumb and the index finger show that the forces are unaffected by the type of food. [1] S. K. Ostwald, M. P. Bernal, S. G. Cron, and K. M. Godwin, “Stress experienced by stroke survivors and spousal caregivers An independent samples t-test has been carried out to during the first year after discharge from inpatient rehabilita- compare the bending angles of the thumb, index finger, and tion,” Topics in Stroke Rehabilitation, vol. 16, no. 2, pp. 93– middle finger and the contact forces exerted by the thumb 104, 2009. and the index finger, respectively, while using different cut- [2] W.-K. Tang, C. G. Lau, V. Mok, G. S. Ungvari, and lery (fork and spoon). The results revealed that the bending K.-S. Wong, “Burden of Chinese stroke family caregivers: the angles of the index finger and the thumb are influenced by Hong Kong experience,” Archives of Physical Medicine and the type of cutlery unlike the middle finger which remains Rehabilitation, vol. 92, no. 9, article 21878218, pp. 1462– unaffected. Thus, we can conclude that the motion of the 1467, 2011. middle finger is not affected by the type of food characteris- [3] K. M. Godwin, S. K. Ostwald, S. G. Cron, and J. Wasserman, tics and the types of cutlery (fork/spoon) being used. The “Long-term health-related quality of life of stroke survivors independent samples t-test also revealed that the contact and their spousal caregivers,” Journal of Neuroscience Nursing, force exerted by the thumb and the index finger is not influ- vol. 45, no. 3, pp. 147–154, 2013. enced by the cutlery. Hence, it can be concluded that the con- [4] W. Duggleby, A. Williams, S. Ghosh et al., “Factors influencing tact force exerted by the index fingertip and the thumb tip are changes in health related quality of life of caregivers of persons not influenced by the different food characteristics nor by the with multiple chronic conditions,” Health and Quality of Life cutlery being used. Outcomes, vol. 14, no. 1, p. 81, 2016. [5] W. Mudzi, A. Stewart, and E. Musenge, “Caregiver strain and associated factors 12 months post stroke: impact of caregiver 5. Conclusion education,” Physiotherapy, vol. 101, no. 1, article e1055, 2015. [6] C. A. Gbiri, O. A. Olawale, and S. O. Isaac, “Stroke manage- In this study, human hand motion analysis was carried out ment: informal caregivers’ burdens and strians of caring for on five different eating activities with six subjects. The stroke survivors,” Annals of Physical and Rehabilitation Medi- ANOVA and t-test results revealed that the bending motion cine, vol. 58, no. 2, pp. 98–103, 2015. of the index finger and the thumb is affected with respect to [7] A. Westergren, I. R. Hallberg, and O. Ohlsson, “Nursing different food characteristics as well as the type of cutlery assessment of dysphagia among patients with stroke,” Scandi- used, that is, a fork and a spoon, whereas the bending motion navian Journal of Caring Sciences, vol. 13, no. 4, pp. 274–282, of the middle finger remains unaffected. In addition, the con- 1999. tact force exerted by the thumb tip and index fingertip [8] A. Westergren, M. Unosson, O. Ohlsson, B. Lorefält, and I. R. remains unaffected with respect to the different types of food Hallberg, “Eating difficulties, assisted eating and nutritional status in elderly (≥65 years) patients in hospital rehabilita- and cutlery used. These results can be useful in the future to tion,” International Journal of Nursing Studies, vol. 39, no. 3, differentiate hand motions dependent on different eating pp. 341–351, 2002. activities and the different cutlery (fork/spoon) used. It can [9] A. Westergren, “Detection of eating difficulties after stroke: a further be used in the development of a mathematical model systematic review,” International Nursing Review, vol. 53, of the hand for eating rehabilitation purposes. no. 2, pp. 143–149, 2006. [10] C. Jacobsson, K. Axelsson, P. O. Osterlind, and A. Norberg, “How people with stroke and healthy older people experience Conflicts of Interest the eating process,” Journal of Clinical Nursing, vol. 9, no. 2, The authors declare that they have no conflicts of interest. pp. 255–264, 2000. 12 Applied Bionics and Biomechanics [11] Z. Ju and H. Liu, “Human hand motion analysis with multi- sensory information,” IEEE/ASME Transactions on Mechatro- nics, vol. 19, no. 2, pp. 456–466, 2014. [12] R. A. R. C. Gopura, K. Kiguchi, and E. Horikawa, “A study on human upper-limb muscles activities during daily upper-limb motions,” International Journal of Bioelectromagnetism, vol. 12, no. 2, pp. 54–61, 2010. [13] X. Tang, Y. Liu, C. Lv, and D. Sun, “Hand motion classification using a multi-channel surface electromyography sensor,” Sen- sors, vol. 12, no. 12, pp. 1130–1147, 2012. [14] J. J. Cabibihan, I. Ahmed, and S. S. Ge, “Force and motion analyses of the human patting gesture for robotic social touch- ing,” in 2011 IEEE 5th International Conference on Cybernetics and Intelligent Systems (CIS), pp. 165–169, Qingdao, China, [15] J. Rosen, J. C. Perry, N. Manning, S. Burns, and B. Hannaford, “The human arm kinematics and dynamics during daily activ- ities - toward a 7 DOF upper limb powered exoskeleton,” ICAR '05 Proceedings, 12th International Conference on Advanced Robotics, 2005, 2005, pp. 532–539, Seattle, WA, USA, July, [16] J. Ah, E. Joo, P. Woo, H. Ram, J. Hyuk, and J. Nam, “Three- dimensional kinematic motion analysis of door handling task in people with mild and moderate stroke,” Physical Therapy Rehabilitation Science, vol. 5, pp. 143–148, 2016. [17] I. Aprile, M. Rabuffetti, L. Padua, E. Di Sipio, C. Simbolotti, and M. Ferrarin, “Kinematic analysis of the upper limb motor strategies in stroke patients as a tool towards advanced neuror- ehabilitation strategies: a preliminary study,” BioMed Research International, vol. 2014, Article ID 636123, 8 pages, 2014. [18] N. H. Adnan, K. Wan, A. B. Shahriman et al., “Measurement of the flexible bending force of the index and middle fingers for virtual interaction,” Procedia Engineering, vol. 41, pp. 388– 394, 2012. [19] N. Adnan, K. Wan, and A. Shahriman, “Accurate measure- ment of the force sensor for intermediate and proximal pha- langes of index finger,” International Journal of Computers Applications, vol. 45, no. 15, pp. 59–65, 2012. [20] A. G. Bluman, Elementary Statistics : A Step by Step Approach, McGraw-Hill, New York, NY, USA, 2012. [21] D. J. Biau, B. M. Jolles, and R. Porcher, “P value and the theory of hypothesis testing: an explanation for new researchers,” Clinical Orthopaedics and Related Research, vol. 468, no. 3, pp. 885–892, 2010. [22] R. A. Fisher, Statistical Methods for Research Workers, Springer, New York, NY, USA, 1992. International Journal of Advances in Rotating Machinery Multimedia Journal of The Scientific Journal of Engineering World Journal Sensors Hindawi Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 http://www www.hindawi.com .hindawi.com V Volume 2018 olume 2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Journal of Control Science and Engineering Advances in Civil Engineering Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com Journal of Journal of Electrical and Computer Robotics Engineering Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 VLSI Design Advances in OptoElectronics International Journal of Modelling & Aerospace International Journal of Simulation Navigation and in Engineering Engineering Observation Hindawi Hindawi Hindawi Hindawi Volume 2018 Volume 2018 Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com www.hindawi.com www.hindawi.com Volume 2018 International Journal of Active and Passive International Journal of Antennas and Advances in Chemical Engineering Propagation Electronic Components Shock and Vibration Acoustics and Vibration Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal

Applied Bionics and BiomechanicsHindawi Publishing Corporation

Published: Feb 4, 2018

References