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

Learn More →

Close2U: An App for Monitoring Cancer Patients with Enriched Information from Interaction Patterns

Close2U: An App for Monitoring Cancer Patients with Enriched Information from Interaction Patterns Hindawi Journal of Healthcare Engineering Volume 2020, Article ID 3057032, 13 pages https://doi.org/10.1155/2020/3057032 Research Article Close2U: An App for Monitoring Cancer Patients with Enriched Information from Interaction Patterns 1 1,2 3,4 1 Javier Navarro-Alama ´ n, Raquel Lacuesta, Iva ´ n Garc´ ıa-Magariño , Jesu ´ s Gallardo, 5 6,7 Elena Ibarz, and Jaime Lloret Department of Computer Science and Engineering of Systems, University of Zaragoza, Zaragoza, Spain Instituto de Investigacio ´n Sanitaria Arago ´n, University of Zaragoza, Zaragoza, Spain Department of Software Engineering and Artificial Intelligence, Complutense University of Madrid, Madrid, Spain Instituto de Tecnolog´ ıa del Conocimiento, UCM, Madrid, Spain Departamento de Ingenier´ ıa Meca ´nica, Universidad de Zaragoza, Zaragoza, Spain Integrated Management Coastal Research Institute, Universitat Politecnica de Valencia, Valencia, Spain School of Computing and Digital Technologies, Staffordshire University, Stoke, UK Correspondence should be addressed to Iva´n Garc´ıa-Magariño; igarciam@ucm.es Received 27 October 2019; Revised 29 May 2020; Accepted 16 June 2020; Published 15 July 2020 Academic Editor: Jesus Fontecha Copyright © 2020 Javier Navarro-Alaman et al. .is 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. .e management of cancer patients’ symptoms in doctor consultations is a cornerstone in clinical care, this process being fundamental for the follow-up of the evolution of these. .is article presents an application that allows collecting periodically and systematically the data of cancer patients and their visualization by the medical team. In this article, we made the analysis, design, implementation, and final evaluation by analyzing the correlation of this data collection with interaction patterns to determine how the user information can be enriched with information from the interaction patterns. We have followed an agile methodology based on the iterative and incremental development of successive prototypes with increased fidelity, where the requirements and solutions have evolved over time according to the need and assessments made. .e comprehensive analysis of the patient’s condition allowed us to perform a first analysis of the correlation of the states of patients concerning mood, sleeping quality, and pain with the interaction patterns. A future goal of this project is to optimize the process of data collection and the analysis of information. Another future goal is to reduce the time dedicated to reporting the evolution of symptoms in face-to-face consultations and to help professionals in analyzing the patient’s evolution even in the period that has not been attended in person. medical fields, such as the app for monitoring patients with 1. Introduction risk of suffering cerebral strokes by means of cloud services .e management of the symptomatology of patients con- [1]. .ere are systems based on wearables that also monitor sumes more than half the time spent by health professionals in patients, like the wireless body area network system for monitoring the patient’s evolution. It is considered a cor- controlling obesity [2]. .ere are proposals for monitoring nerstone of clinical care, particularly for patients with chronic patients with 5G networks, such as the architecture and diseases. However, symptoms and physical disabilities are not protocol for smart eHealth monitoring [3]. detected by health professionals until the patient has an .e number of people who suffer and overcome cancer is appointment. As a result, the opportunity to intervene and continuously increasing, so there is a great concern for their alleviate suffering is lost. In addition, the incomplete infor- quality of life and quality of care. Most of these people adapt mation in the Electronic Health Records (EHRs) limits the well psychologically and physically after completing their ability of medical professionals to understand the results of initial treatment. .us, it is essential to identify the needs in patients. .ere are several mHealth applications that can progress to provide support as soon as possible. .e need of remotely monitor patients to overcome this barrier in many monitoring oncological patients is considered a fundamental 2 Journal of Healthcare Engineering aspect in their follow-up [4], the main objective being the disease. In addition, one can motivate them to complete maximization of the benefits of treatments adjusted to the their daily tasks and their treatment such as daily activities effects and reactions. .e impact of knowing the disease and medication. One of the main objectives is to improve the greatly influences the roles played by patients at home, at monitoring of patients with the aim of also improving the work, and in their community [5]. quality of life of people. For this purpose, the HCI discipline .e articles [6, 7] agree that there are usability problems contributes to the design of usable interfaces that improve in health applications. .erefore, the participation of pa- the record of their evolution. tients in the design is very important, so that it can have an .e analysis of emotions has begun to acquire scientific impact on the state of patients through the use of the app. interest according to the possible applications, increasing the According to [6], patients are interested in using these new number of studies related to the emotional component and technologies, but current tools are not suitable for them. For its form of evaluation in different areas [9]. this reason, we aim at analyzing the data and give patients .e role of emotional communication in the context of control over the collection of their data and see how their HCI is very relevant and challenging when considering the different areas of technological application [10], like in the status affects the use of the application. In this work, the end users are cancer patients, being the example that focuses on remote medical care and assistive evaluation of all the possible influential factors on their technologies. In this area, there is an increasing interest in evolution (psychic, physical, treatments, and pharmacology) this type of research, which includes a synthesis of the effects fundamental. For a better design and development of the and emotion. Emotions affect human behavior and the application, we follow a patient-centered design, with pa- system, so when a user is using an application, their emo- tients and medical professionals in successive interviews tional state can affect the usability of the system [11]. with them. .e term mHealth is defined as the union of mobile .e objective of the present study is to perform an integral computing, medical sensors, and communication technol- follow-up of cancer patients through the use of a mobile ogies for health care [12]. MHealth is an upgrowing field application that allows registering the levels of physical and about the application of mobile technologies in health, which in recent years has emerged as an important segment psychological follow-up, as well as additional influential parameters in its evolution: treatments, medication, activities of telemedicine and its main objective is to improve health carried out, and quality of sleep among others. .is moni- services, integrating the benefits of mobility and ubiquity, toring allowed professionals to evaluate the effectiveness of typical of mobile systems, to the care treatments of tradi- the treatments, as well as to guide or personalize their se- tional health, pretending to bring health care to people and lection, identifying additionally the population at risk and not people to the health system. MHealth applications are thus improving the quality of life of patients. creating mechanisms for the exchange of information re- Additionally, parameters are recorded to take interaction lated to health care, even in remote and low-income areas, patterns into account, such as the choice changes made by due to the large area of coverage and social influence of the patient when replying to each question and the whole mobile telephone networks, becoming a factor strategic to reply time. .rough the study of the collected data, we made save lives [13]. In this context of mHealth, we will focus on an analysis of the relationship between user variables the follow-up of cancer patients. (concerning mood, sleeping quality, and pain) and the in- In addition to the intervention of the doctors in the teraction patterns. consultations with their patients, another relevant aspect is .e rest of the paper is structured as follows. .e related the optimization of their next consultations before doing work is studied in Section 2. Section 3 analyzes the materials them. In this way, they can improve the way in which the and methods that we use in this research. .e results of the medical professionals dictate a new treatment. .ere is an research are presented in Section 4. Section 5 discusses the application of medical treatment to optimize the consulta- results and the limitations. Finally, the conclusion and future tions of medical patients according to the known preferences work are mentioned in Section 6. and other selection criteria [14]. .at work refers to opti- mizing the appointments of the patients, the insurance payment options, the treatment in the medical facilities, and 2. Related Work other aspects of the interactions between patients. .e ap- .e Human-Computer Interaction (HCI) applied to the plication gathers data from the patient’s history and cal- culates a placement score in a medical center based on a field of health is increasingly essential in the development of interactive systems for medical professionals and patients plurality of parameters of the user’s medical history asso- ciated with the user’s medical history and schedules the [8]. .e work of medical professionals (doctors, psycholo- gists, etc.) can be simplified by the continuous use of medical care consultation at a medical center that provides technology to obtain information in real time from patients. the highest placement score in a medical center. Many more Patients can be supported by technology during their applications are being developed and can be downloaded treatment or recovery process, both in the diagnosis and through the Google Play Store for Android phone users. treatment phases, and in the case of having successfully Some of the mHealth systems are for some hereditary overcome the disease. Providing patients with these appli- diseases and disorders [15], cancer-based apps [16–19], or studies on the impact of stress or emotions on the interaction cations can be beneficial for them as they are more pleasant, or at least not as monotonous, in their treatment of the with the application [11, 20]. Journal of Healthcare Engineering 3 Android Mobile Informatics Application for Some contemplates the integral monitoring of patients. Patients Hereditary Diseases and Disorders (AMAHD) is a com- were able to see and manage both their medication and their plementary framework for medical practitioners and pa- appointments. For the part of the appointments that have tients. .e mobile application will help to sensitize and both with the psychologist and with the doctor, a section was complement the efforts of biomedical, medical, and bio- designed in which they can visualize the appointments in a informatics researchers working in the areas of inheritance calendar or in a detailed list, and they can be managed by research and genetics. AMAHD has proven to be a valuable them. .e part of the patient’s medication has also been resource for the research company in the battle against designed, in this way they can keep track of his prescriptions, hereditary diseases and disorders [15]. allowing them to visualize and manage them themselves; the medication can be viewed weekly or through a list with the .ere are some cancer-based applications. For instance, details. a smartphone-based pain management app for adolescents .e methodology used for the application Close2U with cancer provides patients pain management support was an agile methodology based on the iterative and based on their individual pain [16]. Another article evaluates incremental development of successive prototypes with the usability of their app and shows the results of the iterative increased functionality. .e application has evolved in its development of their app. Its authors inform other devel- phases of analysis, design, implementation, and evalua- opers and researchers in development, integration, and tion. .e successive refinements have allowed the evo- evaluation of mobile health apps and services that support lution of the prototypes and the increase of their cancer patients in managing their health-related issues [17]. functionality and quality. .e agile methodology may be .e goal of another app is to stabilize a daily functional the most appropriate for projects that suffer a high activity in breast cancer patients. App-using participants number of changes and need more control and com- could more frequently report adverse events, and those munication with the client in real time and allows both under supervision made fewer and more precise entries than adapting to problems that may arise and making the unsupervised participants [18]. .e last reviewed app is a necessary changes at the beginning of each phase, without smartphone app framework for segmented cancer care having to wait to finish all the actions. coordination, which provides both medical risk assessment Nowadays, it is crucial to incorporate the user into the and health care monitoring functions [17]. However, none of design and implementation cycle of a mobile application. In these apps considered information from interaction patterns addition, in the medical-scientific research community, to enrich the extracted information, as the current work there is a growing thought that states that the control and does. prevention of cancer must incorporate a communication Besides implemented applications, some papers have with the patient [10]. In this way, the benefit of current studied the impact of emotions over the medical applications medical discoveries in diagnosis and treatment is maxi- and how emotions can be related to the usability of the mized, particularly in the emerging era of personalized system that they are interacting with [11, 20]. Defining the medicine. Although patient-medical communication has stress level of the user can train the system in such a way that focused on results such as patient satisfaction, under- it could not only detect the user’s stress level but can also standing, and assessment, we must strengthen the under- modify itself accordingly, thereby increasing the usability of standing of how these can impact on their attitude and the the system and the user satisfaction [11]. In [20], it is shown results of the disease [10]. .roughout all phases of devel- that experiencing negative emotions during the use of the opment, we have worked in cooperation with the psy- system can negatively influence important user behaviors, chologists of the AECC (Spanish Association against including the client’s decisions regarding the application. Cancer) in Teruel, who were present throughout the cycle of In conclusion, monitoring and reminder applications design, implementation, and evaluation of the product. have been helpful for users in assisting them in monitoring To design and develop the application, a conceptual and keep track of the patient’s health care records as well as framework was first made; then the necessary algorithms their medication intake, and there are some studies that for data capture were designed and the software tools show how the emotions can affect the usability of mobile necessary to design it were selected. .e developed ap- applications. .is motivates the development of Close2U plication allows one to treat and use all the information application with the intention to help the medical practi- collected about physical condition, activities, treatments, tioners with the analysis that the app can bring them and and so on. Each element of the follow-up was funda- help them in improving the treatment for the patient. In mental to know the overall situation of the patient, addition, a novelty of Close2U app is that it shows that the maintaining their privacy. Information was collected and, analysis of interaction patterns can enrich self-reported user in order of priority, actions to be carried out were defined. information. .is work analyzes the correlations among cancer-patient variables and interaction-pattern variables. 3.1.1. User Interface of Close2U App. .is section presents the user interface showing the most relevant screens of the 3. Materials and Methods final implementation of Close2U app. .e changes that 3.1. Materials: Close2U App. .e main material for this occurred in the successive evaluations will be explained, research is the mobile application Close2U. .e application since changes were made throughout the implementation, 4 Journal of Healthcare Engineering (a) (b) Figure 1: Screens of the user interface 1: (a) mood selection and (b) sleep quality selection. with the help of the evaluation we made together with the Another valued aspect was the parts where the patient psychologists and the cancer patients. felt pain or discomfort, and a screen was made with which Figure 1(a) shows the screen in which the patient is asked the patient could visualize a body in which a pain could be indicated by buttons to make it more visual. .is is about their mood. It is a question that had many changes since at the beginning, the emotions were not taken into account, intended to be more intuitive for the patient as seen in Figure 2(c), since at first there was no image but simply only the mood, which limited the patient to specifying his mood. Once the emotions were introduced, it was possible to buttons. clarify in more detail how they feel, opening the way to a new “mood register” called “Undecided”; this state of mind de- 3.1.2. Internal Functioning of Close2U App. .e purpose of pends on the emotions that the patient selects. When the the design phase was to ensure that the developed appli- patient selects various moods, the medical professional can cation meets the requirements of the end user before the see the situation of patient’s indecision. prototype is translated into the production application. We In the screen of Figure 1(b), one can see the question used the Unified Modeling Language (UML) for defining the about sleep quality. .is screen was modified several times design diagrams. To carry out the plan of user activities, the psychol- according to the requirements and evaluation meetings established by the psychologists of the AECC Teruel; it was ogists of the AECC Teruel were taken into consideration. decided to show the patient in a numerical form from 0 to In this way, the interactions that the user must perform 10, since it is more comfortable and clearer in terms of with the application were identified. Figure 3 shows how usability. .is same change was also made with the level of patient interacts with the application when carrying out pain, which is presented later in Figure 2(b). the surveys. In each question, the app checks that the Figure 2(a) shows the options to assess whether you feel answer is valid and continues to the next question until pain or discomfort, or do not feel both. Initially only the pain the end of the questionnaire. .e replies are sent through was registered, but later considering the comments from the API, and it will be verified that the survey has been patients and psychologists, it was later decided to add a registered. In the user interface, the app confirms the differentiation between pain and discomfort. patient that the survey has been successfully registered. Journal of Healthcare Engineering 5 (a) (b) (c) Figure 2: Screens of the user interface 2: (a) pain/discomfort selection, (b) pain level selection, and (c) body parts with pain selection. with its emulator to perform the patient part. .e imple- For the selection of the questions to be shown in the survey, a previous analysis was carried out and they were mentation and validation were carried out over twelve refined through successive evaluations. Next, the interface months. associated with each of them was designed, as described in .e validation of the application has been made from the Section 3.1.1. For this stage, the flow of activities and their requirements analysis phases until the final implementation order were designed, so that in this way the patient will with the medical professionals of the AECC Teruel. .e first “connect” better with the survey and neither feels bad tests with patients allowed us to correct the existing con- when doing it nor gets bored along with it. ceptual and design errors. In the next stage of the design, once the activities were Once the first functional prototype of the application defined, a sequence diagram was defined, and Figure 4 presents was ready, tests were carried out with the users. Based on this diagram, in which we show how the surveys work internally them, we got back to the design phase and improved the design. .rough the evaluation carried out, the interfaces in the application. For all the questions shown on the screen, in parallel when the patient answers each question, the change is were modified to improve aspects of the usability and added to that answer. .en, when the patient confirms their functional application, for example, so that this would be answer, it goes to the next question. For the first and last intuitive for the patient. In addition, being able to assess the questions, there are special sequences. In the first question, the mood of the users during the use of the live system had a start time of the survey is recorded and in the last question the practical meaning for the design and improvement of our total time is calculated, the response is sent to the server, and application. finally, the survey ends. At the sequence diagram, the objects are shown as 3.2. Methods lifelines along the survey and with their interactions over time represented as arrows from the origin lifeline to the 3.2.1. Participants. .ere were 23 users that voluntarily participated in this study, without any economical com- destination lifeline. Sequence diagrams are appropriate for showing which objects communicate with each other and pensation. .ey were 50.21 years old in average with a the messages that trigger those communications. standard deviation (SD) of 10.09. .ey were 6 males and 17 For the implementation, the IDE (Integrated Devel- females. .e sample included 7 cancer patients and 16 opment Environment) Android Studio was used, together healthy people. 6 Journal of Healthcare Engineering Question Web service Yes Patient Is the last question? Show question No First question setInitTime() Show question? Parallel setAnswer () Select answers addAnswerChange () check (RegSurvey) Yes Valid answers? goNext (RegSurvey) No Last question surveyTime () Show error post (RegSurvey) send and receive API data Next question goFinish (RegSurvey) Patient Question Web service Figure 4: Sequence diagram of Close2U app. Send data to the API (i) Category 1: overwhelmed, agitated, ashamed, de- pressed, angry, exhausted, hurt, scared, alone, ap- No e survey has been registered? prehensive, defeated, desperate, despondent, exhausted, helpless, angry, frustrated, impatient, pessimistic, self-critical, irritated, defensive, de- Show survey done spised, and resentful (ii) Category 2: guilty, stressed, angry, bad, grumpy, dizzy, nervous, sad, discouraged, apathetic, worried, anxious, agitated, hurt, disappointed, distressed, Figure 3: Activity diagrams of Close2U app. nostalgic, offended, hungry, lacking confidence, impotent, repentant, insecure, and rejected 3.2.2. Measures. We evaluated sleep quality with the self- (iii) Category 3: bored, confused, asleep, busy, pensive, reported question “How did you sleep?” replied in a 0–10 tired, neutral, unequal, confused, hesitant, lazy, disconnected, reserved, indifferent, apprehensive, range (zero meaning the worst sleep quality and ten meaning the best). We determined the existence of pain conflicted, disconnected, reserved, indifferent, scattered, restless, sensitive, and vulnerable with the question “Have you noticed any pain or dis- comfort in the last 12 hours?” with an answer from “yes,” (iv) Category 4: well, happy, awake, relieved, cared, “no,” and “discomfort” options. If the user replied affir- empathetic, not critical, confident, calm, and sincere matively, then the app asked, “What level of pain/dis- (v) Category 5: super, proud, satisfied, confident, alive, comfort do you have now?” which was answered in a 1–10 enthusiastic, strong, encouraged, excited, grateful, range (one, minimum pain level and ten, maximum). .e hopeful, and open-minded app recorded zero level if the user had previously an- swered to feel neither pain nor discomfort. We evaluated We labeled the responses with more than one mood as the mood by asking the user “How are you feeling?” with a undecided. We assessed the indecision for each question multiple response among 97 moods classified in five mood with the number of times that the user changes their reply categories (1: horrible, 2: discouraged, 3: normal, 4: an- before submitting it. We applied this measure to (a) sleep imated, and 5: radiant), categorized by the medical pro- quality, (b) whether the user felt pain, (c) the level of this fessionals with who we worked in every step of the design pain, and (d) the mood. and development of the app. .ese moods were the fol- We used System Usability Scale (SUS) [21] to measure lowing ones: the usability of the app. Journal of Healthcare Engineering 7 Table 1: Relation of mood changes and time of realization. Average Undecided Radiant Animated Normal Discouraged Horrible Num. changes when selecting mood (#) 1.57 0.22 0.39 0.24 0.67 1.50 Survey completion time (mm:ss) 00:57 01:08 00:57 01:00 01:16 01:36 2.00 04:00 1.75 03:30 1.50 03:00 1.25 02:30 1.00 02:00 0.75 01:30 0.50 01:00 0.25 00:30 0.00 00:00 Undecided Radiant Animated Normal Discouraged Horrible Mood Time Changes Figure 5: Relation of mood changes and time of realization. 3.2.3. Protocol. Firstly, we followed a user-centered design. usually take more changes before selecting their final state. .ree psychologists familiar with cancer tested the Close2U Also, in this case, they select more than one state of mind. app during the whole development process and provided .e average of the number of changes in this case is 1.57. feedback. .e developer improved the app following their Regarding the total time of the survey, this was generally recommendations until they were satisfied, to achieve a high longer when the patient felt worse. level of usability. In this way, the functionalities of the app In a first approximation, it can be estimated that the were enriched according to the common specific needs of indecision corresponds to positive moods, since the average cancer patients and their doctors. completion time of the undecided survey was similar to the .e app was uploaded to Google Play, which is the main times of positive states. store of apps for Android. We encouraged cancer patients to Table 2 shows the average of the changes made by the use the app in a noncontrolled environment, advising them patient each time he had to choose the level of pain and the to use the app regularly. time in which he performs the survey. We measured the collected information about users What we can observe in the graph of Figure 6 is (related to mood, sleep, and pain) and the information from an increase in the number of changes that the patient made interaction patterns (number of changes when selecting when they had more pain, as in the time of conducting the replies to specific questions and the global survey time). We survey that is higher when the patient chose a higher level of analyzed the correlations among all these variables. pain. .is relation was also observed in other experiments, as We measured the usability of the app at the end of the one can see in the next presented graph. study with SUS scale. Table 3 shows an average of the changes made by the patient when he had to decide if he had pain or not, and in 4. Results the case of having pain if it is discomfort or pain and the time in which he performed the survey. Table 1 shows (a) the average of the changes made by the In the graph of Figure 7, it can be seen to what extent it patient each time he had to choose the emotions that mark affected that the patient did not have pain, and we found that his mood and (b) the time the survey took. it took less than half the time to complete the survey. When In Figure 5, it can be observed that the patient modifies comparing pain and discomfort, there were not such dif- his response in a greater number of occasions when his ferences in the time to carry out the survey. mood is negative. When the users are undecided, they Num. changes when selecting mood (#) Survey completion time (mm:ss) 8 Journal of Healthcare Engineering Table 2: Relation of pain level changes and time of realization. Average 1 2 3 4 5 6 7 8 9 10 Num. changes when selecting pain level (#) 0.08 0.42 0.78 0.42 0.93 0.67 0.31 0.61 1.00 1.43 Survey completion time (mm:ss) 01:40 01:29 01:19 01:30 01:16 01:17 00:59 01:43 03:35 02:30 2.00 04:00 1.75 03:30 1.50 03:00 1.25 02:30 1.00 02:00 0.75 01:30 0.50 01:00 0.25 00:30 0.00 00:00 123456789 10 Pain level Time Changes Figure 6: Relation of pain level changes and time of realization. Table 3: Relation of pain changes and time of realization. Average No pain Pain Discomfort Num. changes when selecting pain (#) 0.00 0.04 0.13 Survey completion time (mm:ss) 00:49 01:22 01:24 Table 4 shows an average of the changes that the patient not know whether the correlations would be direct or inverse makes when deciding the level of sleep and the time in which beforehand. the survey is performed, for each sleeping-quality level. Regarding the user variables and interaction patterns Figure 8 shows this information graphically. One can ob- concerning the numbers of changes, the test found significant serve that with the exception of a peak in level 1, the time of correlations in the pairs (a) mood and number of changes when selecting mood, (b) sleep and number of changes when conducting the survey barely changed. By contrast, the number of changes was higher when the patient had slept selecting sleep, (c) pain and number of changes when selecting better. pain, (d) pain and number of changes when selecting pain level, What we can observe in the graph of Figure 8 is that, with (e) pain level and number of changes when selecting mood, (f) the exception of the peak that exists in level 1 of sleep, the pain level and number of changes when selecting sleep, (g) pain time of accomplishment of the survey hardly changed. On level and number of changes when selecting pain, and (h) pain the other hand, the number of changes increased as the level and number of changes when selecting pain level. patient had slept better. Concerning the user variables among each other, the test found In order to determine if there were statistically signifi- significant correlations in all the possible pairs among the user cant correlations, we conducted the Pearson correlation test variables (1) mood, (2) sleep, (3) pain, and (4) pain level. All the between user variables (i.e., mood, sleeping quality, the significant correlations between mood-related variables and other variables had a negative correlation coefficient. However, existence of pain, and pain level) and the number of changes when selecting these variables. Table 5 shows the results of all the significant correlations among non-mood-related var- this test. We considered 497 cases (considering a case each iables had a positive correlation coefficient. time a user performed the complete survey in the app) for In order to determine if the correlations between this analysis. We considered 2-tailed significances as we did survey time and user variables were significant, we Num. changes when selecting pain level (#) Survey completion time (mm:ss) Journal of Healthcare Engineering 9 2.00 04:00 1.75 03:30 1.50 03:00 1.25 02:30 1.00 02:00 0.75 01:30 0.50 01:00 0.25 00:30 0.00 00:00 No pain Pain Discomfort Type of pain Time Changes Figure 7: Relation of pain changes and time of realization. Table 4: Relation of sleep quality changes and time of realization. Sleeping quality 0 1 2 3 4 5 6 7 8 9 10 Num. changes when selecting sleep quality (#) 0.00 0.00 0.14 0.27 0.15 0.23 0.12 0.56 0.75 0.58 1.07 Survey completion time (mm:ss) 01:06 04:25 01:19 00:51 00:53 01:05 00:54 01:07 01:11 00:57 01:05 2.00 05:00 1.75 04:00 1.50 1.25 03:00 1.00 02:00 0.75 0.50 01:00 0.25 0.00 00:00 0 1 23456789 10 Sleep quality Time Changes Figure 8: Relation of sleep quality changes and time of realization. conducted Pearson’s correlation test, and Table 6 shows Regarding usability, the average result of SUS test was the results. .is test found significant correlations be- 69.2 in the standard range of 0–100, and the standard tween (a) survey time and (b) the pain-related variables deviation (SD) was 20.0. Figure 9 shows the results of the including pain and pain level. individual items from the SUS scale. Num. changes when selecting sleep quality (#) Num. changes when selecting pain (#) Survey completion time (mm:ss) Survey completion time (mm:ss) 10 Journal of Healthcare Engineering Table 5: Pearson’s correlation test between user variables and number of changes when selecting the response for these variables, respectively. Num. Num. Num. Num. changes changes changes when changes Pain when selecting Mood Sleep when Pain selecting when level mood pain selecting mood selecting pain level sleep Pearson’s ∗∗ ∗∗ ∗∗ ∗∗ 1 −0.172 −0.133 −0.054 −0.158 −0.080 −0.335 −0.035 correlation Mood Sig. 0.000 0.003 0.234 0.000 0.074 0.000 0.438 (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗ −0.0172 1 0.084 0.166 −0.052 0.023 0.115 0.035 Num. changes correlation when selecting Sig. 0.000 0.061 0.000 0.244 0.608 0.010 0.433 mood (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗∗ ∗∗ −0.133 0.084 1 0.300 0.121 0.027 0.125 0.037 correlation Sleep Sig. 0.003 0.061 0.000 0.007 0.548 0.005 0.412 (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗∗ ∗∗ −0.054 0.166 0.300 1 0.053 −0.025 0.139 0.402 Num. changes correlation when selecting Sig. 0.234 0.000 0.000 0.234 0.581 0.002 0.000 sleep (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗∗ ∗∗ ∗∗ −0.158 −0.052 0.121 0.053 1 0.194 0.703 0.393 correlation Pain Sig. 0.000 0.244 0.007 0.234 0.000 0.000 0.000 (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ −0.080 0.023 0.027 −0.025 0.194 1 0.149 0.049 Num. changes correlation when selecting Sig. 0.074 0.608 0.548 0.581 0.000 0.001 0.274 pain (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗ ∗∗ ∗∗ ∗∗ ∗∗ ∗∗ −0.335 0.115 0.125 0.139 0.703 0.149 1 0.371 correlation Pain level Sig. 0.000 0.010 0.005 0.002 0.000 0.001 0.000 (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗∗ Num. changes −0.035 0.035 0.037 0.402 0.393 0.049 0.371 1 correlation when selecting Sig. mood pain 0.438 0.433 0.412 0.000 0.000 0.274 0.000 (2-tailed) level N 497 497 497 497 497 497 497 497 ∗∗ ∗ Correlation is significant at the 0.01 level (2-tailed). Correlation is significant at the 0.05 level (2-tailed). Table 6: Pearson’s correlation test between survey time and user variables. Survey time Mood Sleep Pain Pain level ∗∗ ∗∗ Pearson’s correlation 1 −0.049 0.007 0.335 0.323 Survey time Sig. (2-tailed) 0.272 0.874 0.000 0.000 N 497 497 497 497 497 ∗∗ ∗∗ ∗∗ Pearson’s correlation −0.049 1 −0.133 −0.158 −0.335 Mood Sig. (2-tailed) 0.272 0.003 0.000 0.000 N 497 497 497 497 497 Journal of Healthcare Engineering 11 Table 6: Continued. Survey time Mood Sleep Pain Pain level ∗∗ ∗∗ ∗∗ Pearson’s correlation 0.007 −0.133 1 0.121 0.125 Sleep Sig. (2-tailed) 0.874 0.003 0.007 0.005 N 497 497 497 497 497 ∗∗ ∗∗ ∗∗ ∗∗ Pearson’s correlation 0.335 −0.158 0.121 1 0.703 Pain Sig. (2-tailed) 0.000 0.000 0.007 0.000 N 497 497 497 497 497 ∗∗ ∗∗ ∗∗ ∗∗ Pearson’s correlation 0.323 −0.335 0.125 0.703 1 Pain level Sig. (2-tailed) 0.000 0.000 0.005 0.000 N 497 497 497 497 497 ∗∗ Correlation is significant at the 0.01 level (2-tailed). 5 In the significant correlations, the sign of correlation co- efficients showed that mood-related variables were inversely correlated with sleep-related variables and pain-related vari- ables. .e inverse correlation between mood and sleep variables makes sense because the pain negatively influenced on the mood of users, as the pain probably did not let them sleep well. .e correlation between the survey time and the pain 2 (including the existence and its level) revealed that the in- teraction patterns also provided useful information about pain. .e survey of the app dedicated a large part to the pain, since the user was asked about the specific locations of the pain in a silhouette-based interface. Users spent more time depending on when they had a higher level of pain. .us, 123456789 10 future applications for cancer patients with silhouette-based Question numbers of SUS interfaces could indirectly help in estimating the pain level. Figure 9: Results of the SUS scale. One of the limitations of this research is the small sample size of participants. .is sample size only allowed us to detect correlations with large effect sizes. .us, we may have not detected correlations with small or medium effect sizes. 5. Discussion We detected a correlation between mood and the number of 6. Conclusion and Future Work changes when selecting the answers. .is reveals that the number of changes when selecting mood is relevant as it is A mobile application has been developed that performs a correlated with useful information. .is implies that cancer complete follow-up of the patient. .is article has focused on apps can collect enriched information from users by counting the self-reported moods, sleeping quality, and pain when the the number of changes when selecting a response in questions. patient is using our application, to provide a tool for psy- .erefore, interaction patterns provided relevant information chologists to (a) improve their treatment, (b) improve ap- concerning moods. pointments with them, (c) carry out a daily monitoring of .e results revealed that users were doubtful (i.e., they patients, and (c) be able to communicate with them outside changed the replies more times) when selecting (a) negative face-to-face consultations. .is increased the patients’ sense moods, (b) high-quality levels of sleep, and (c) high levels of of being taken care of daily, since they did not feel alone. pain. Notice that we distinguished between low and high .is article shows (1) how a mobile health application values for each user variable, and this was concluded from can assist patients in being more active in managing their the sign of correlation coefficients. For example, in the case care, (2) how the emotions affect when completing the of moods, this may reveal that users perceived negative survey, and (3) taking more time or making more changes emotions as a combination of some basic emotions, in ac- usually depend on their mood, pain, or sleep quality at that cordance to the theory of Ekman [22] about basic emotions, time. Our research revealed that interaction-patterns vari- rather than for positive emotions. ables provided useful information from some user variables, Another relevant finding was that pain level was sig- proved with the statistically significant correlations detected nificantly correlated with all the other studied variables in this study. including the user variables and the numbers of changes .is work motivates future research concerning the when selecting any user variable. Hence, pain level was the extraction of implicit information from interaction pat- variable with most relevant information due to its corre- terns in cancer monitoring apps. .e correlation between lations with all the other variables. user variables and interaction patterns triggers a Response in 5-point Likert scale 12 Journal of Healthcare Engineering [2] S. S. MS Mohammed, J. Lloret, and I. Bosch, “Systems and promising research line about getting patients’ informa- WBANs for controlling obesity,” Journal of Healthcare En- tion from their patterns when interacting with apps, being gineering, vol. 2018, Article ID 1564748, 21 pages, 2018. able to enrich the collected information with this non- [3] J. Lloret, L. Parra, M. Taha, and J. Tomas, ´ “An architecture and self-reported information. .is research could also lead to protocol for smart continuous eHealth monitoring using 5G,” take implicit-extracted mood to adapt the interface to the Computer Networks, vol. 129, pp. 340–351, 2017. needs of patients. [4] S. Loibl and B. Lederer, “.e importance of supportive care in According to medical professionals, the app helped them to breast cancer patients,” Breast Care, vol. 9, no. 4, pp. 230-231, collect relevant data about the disease more easily than if they had had to collect these manually. In the future, we plan to [5] F. Cardoso, N. Bese, S. R. Distelhorst et al., “Supportive care analyze how treatment improves and how waiting times are during treatment for breast cancer: resource allocations in reduced before face-to-face consultations. low- and middle-income countries. A Breast Health Global In the future, it is intended to perform the noninvasive Initiative 2013 consensus statement,” 9e Breast, vol. 22, no. 5, pp. 593–605, 2013. recording and measurement of the evolution of patients, by [6] U. Sarkar, G. I. Gourley, C. R. Lyles et al., “Usability of means of devices or sensors, to improve both their treatment commercially available mobile applications for diverse pa- and their quality of life. For this purpose, we will search tients,” Journal of General Internal Medicine, vol. 31, no. 12, available devices or sensors that will support the capture of pp. 1417–1426, 2016. evaluable parameters that affect patients, at the same time [7] H. Fu, S. K. McMahon, C. R. Gross, T. J. Adam, and that our application will collect data. Initially, smart sensors J. F. Wyman, “Usability and clinical efficacy of diabetes mobile and wristbands will be used to record the heart rate and the applications for adults with type 2 diabetes: a systematic Electrocardiograms (ECG), blood pressure, and Heart Rate review,” Diabetes Research and Clinical Practice, vol. 131, Variability (HRV). We will also explore the registry of other pp. 70–81, 2017. parameters such as oxygen level in the blood, skin and body [8] D. DAuria, F. Persia, and B. Siciliano, “Human-computer temperature, and breathing patterns among others. interaction in healthcare: how to support patients during their wrist rehabilitation,” in Proceedings of the 2016 IEEE Tenth International Conference on Semantic Computing (ICSC), Data Availability pp. 325–328, IEEE, Laguna Hills, CA, USA, 2016 February. [9] G. Sacca Espinoza, “Estrategia para el reconocimiento de .e data are not available to avoid privacy infringement. emociones colectivas en entornos sociales basados en tic strategy for recognizing collective emotions in social environments based on ict,” M.S. thesis, Universidad Autonoma ´ de Madrid, Conflicts of Interest Madrid, Spain, 2015. [10] A. Esposito, A. M. Esposito, and C. Vogel, “Needs and .e authors declare that there are no conflicts of interest challenges in human computer interaction for processing regarding the publication of this paper. social emotional information,” Pattern Recognition Letters, vol. 66, pp. 41–51, 2015. [11] S. Saxena, P. Saxena, and S. K. Dubey, “Various levels of Acknowledgments human stress & their impact on human computer interac- tion,” in Proceedings of the 2013 International Conference on .e authors would like to thank the psychologists of the Human Computer Interactions (ICHCI), pp. 1–6, IEEE, AECC of Teruel for their willingness to help them in im- Chennai, India, August 2013. proving the application. .is work was funded by the re- [12] R. S. H. Istepanian, E. Jovanov, and Y. T. Zhang, “Guest search projects “New Technologies in Cancer Monitoring editorial introduction to the special section on m-health: and Treatment” and “Advances in Research, Diagnosis and beyond seamless mobility and global wireless health-care Monitoring of Bone Cancer Patients as a Primary or Sec- connectivity,” IEEE Transactions on Information Technology ondary Tumor” by the University Foundation “Antonio in Biomedicine, vol. 8, no. 4, pp. 405–414, 2004. Gargallo.” .is work was partly funded by the Spanish [13] V. W. Consulting, mHealth for Development: 9e Opportunity Science, Innovation and University Ministry (MCIU), the of Mobile Technology for Healthcare in the Developing World, UN Foundation-Vodafone Foundation Partnership, Wash- National Research Agency (AEI), and the EU (FEDER) ington DC, USA, 2009. through the contract RTI2018-096986-B-C31 and by the [14] D. R. Albro, K. A. Darby, R. J. Dudzinski et al., “Medical Aragonese Government and the EU through the FEDER treatment application for optimizing medical patient visits 2014–2020 “Construyendo Europa desde Aragon” ´ action based on known preferences and other selection criteria,” U.S. (Group T25_17D). .ey also acknowledge the partial sup- Patent Application No. 10262384, JUSTIA Patent, Mountain port from CYTED (ref. 518RT0558) and MCIU (TIN2017- View, CA, USA, 2019. 88327-R). [15] O. Oluwagbemi, F. Oluwagbemi, and C. Ughamadu, “An- droid mobile Informatics application for some hereditary diseases and disorders (AMAHD): a complementary frame- References work for medical practitioners and patients,” Informatics in Medicine Unlocked, vol. 2, pp. 38–69, 2016. [1] L. Garc´ıa, J. Tomas, ´ L. Parra, and J. Lloret, “An m-health application for cerebral stroke detection and monitoring [16] L. A. Jibb, B. J. Stevens, P. C. Nathan, E. Seto, J. A. Cafazzo, and J. N. Stinson, “A smartphone-based pain management using cloud services,” International Journal of Information Management, vol. 45, pp. 319–327, 2019. app for adolescents with cancer: establishing system Journal of Healthcare Engineering 13 requirements and a pain care algorithm based on literature review, interviews, and consensus,” JMIR Research Protocols, vol. 3, no. 1, p. e15, 2014. [17] J. Mirkovic, D. R. Kaufman, and C. M. Ruland, “Supporting cancer patients in illness management: usability evaluation of a mobile app,” JMIR mHealth and uHealth, vol. 2, no. 3, p. e33, [18] M. Egbring, E. Far, M. Roos et al., “A mobile app to stabilize daily functional activity of breast cancer patients in collab- oration with the physician: a randomized controlled clinical trial,” Journal of Medical Internet Research, vol. 18, no. 9, p. e238, 2016. [19] T. He, R. Ogunti, M. Puppala et al., “A smartphone app framework for segmented cancer care coordination,” in Proceedings of the 2016 IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI), pp. 372–375, IEEE, Las Vegas, NV, USA, February 2016. [20] M. Hibbeln, T. Martin, J. L. Jenkins et al., “How is your user feeling? Inferring emotion through human-computer inter- action devices,” MIS Quarterly, vol. 41, no. 1, pp. 1–21, 2017. [21] J. Brooke, “SUS-A quick and dirty usability scale,” Usability Evaluation in Industry, vol. 189, no. 194, pp. 4–7, 1996. [22] P. Ekman, “An argument for basic emotions,” Cognition & Emotion, vol. 6, no. 3-4, pp. 169–200, 1992. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Healthcare Engineering Hindawi Publishing Corporation

Close2U: An App for Monitoring Cancer Patients with Enriched Information from Interaction Patterns

Loading next page...
 
/lp/hindawi-publishing-corporation/close2u-an-app-for-monitoring-cancer-patients-with-enriched-Cco8O2Y0xi

References (23)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2020 Javier Navarro-Alamán 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
2040-2295
eISSN
2040-2309
DOI
10.1155/2020/3057032
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Healthcare Engineering Volume 2020, Article ID 3057032, 13 pages https://doi.org/10.1155/2020/3057032 Research Article Close2U: An App for Monitoring Cancer Patients with Enriched Information from Interaction Patterns 1 1,2 3,4 1 Javier Navarro-Alama ´ n, Raquel Lacuesta, Iva ´ n Garc´ ıa-Magariño , Jesu ´ s Gallardo, 5 6,7 Elena Ibarz, and Jaime Lloret Department of Computer Science and Engineering of Systems, University of Zaragoza, Zaragoza, Spain Instituto de Investigacio ´n Sanitaria Arago ´n, University of Zaragoza, Zaragoza, Spain Department of Software Engineering and Artificial Intelligence, Complutense University of Madrid, Madrid, Spain Instituto de Tecnolog´ ıa del Conocimiento, UCM, Madrid, Spain Departamento de Ingenier´ ıa Meca ´nica, Universidad de Zaragoza, Zaragoza, Spain Integrated Management Coastal Research Institute, Universitat Politecnica de Valencia, Valencia, Spain School of Computing and Digital Technologies, Staffordshire University, Stoke, UK Correspondence should be addressed to Iva´n Garc´ıa-Magariño; igarciam@ucm.es Received 27 October 2019; Revised 29 May 2020; Accepted 16 June 2020; Published 15 July 2020 Academic Editor: Jesus Fontecha Copyright © 2020 Javier Navarro-Alaman et al. .is 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. .e management of cancer patients’ symptoms in doctor consultations is a cornerstone in clinical care, this process being fundamental for the follow-up of the evolution of these. .is article presents an application that allows collecting periodically and systematically the data of cancer patients and their visualization by the medical team. In this article, we made the analysis, design, implementation, and final evaluation by analyzing the correlation of this data collection with interaction patterns to determine how the user information can be enriched with information from the interaction patterns. We have followed an agile methodology based on the iterative and incremental development of successive prototypes with increased fidelity, where the requirements and solutions have evolved over time according to the need and assessments made. .e comprehensive analysis of the patient’s condition allowed us to perform a first analysis of the correlation of the states of patients concerning mood, sleeping quality, and pain with the interaction patterns. A future goal of this project is to optimize the process of data collection and the analysis of information. Another future goal is to reduce the time dedicated to reporting the evolution of symptoms in face-to-face consultations and to help professionals in analyzing the patient’s evolution even in the period that has not been attended in person. medical fields, such as the app for monitoring patients with 1. Introduction risk of suffering cerebral strokes by means of cloud services .e management of the symptomatology of patients con- [1]. .ere are systems based on wearables that also monitor sumes more than half the time spent by health professionals in patients, like the wireless body area network system for monitoring the patient’s evolution. It is considered a cor- controlling obesity [2]. .ere are proposals for monitoring nerstone of clinical care, particularly for patients with chronic patients with 5G networks, such as the architecture and diseases. However, symptoms and physical disabilities are not protocol for smart eHealth monitoring [3]. detected by health professionals until the patient has an .e number of people who suffer and overcome cancer is appointment. As a result, the opportunity to intervene and continuously increasing, so there is a great concern for their alleviate suffering is lost. In addition, the incomplete infor- quality of life and quality of care. Most of these people adapt mation in the Electronic Health Records (EHRs) limits the well psychologically and physically after completing their ability of medical professionals to understand the results of initial treatment. .us, it is essential to identify the needs in patients. .ere are several mHealth applications that can progress to provide support as soon as possible. .e need of remotely monitor patients to overcome this barrier in many monitoring oncological patients is considered a fundamental 2 Journal of Healthcare Engineering aspect in their follow-up [4], the main objective being the disease. In addition, one can motivate them to complete maximization of the benefits of treatments adjusted to the their daily tasks and their treatment such as daily activities effects and reactions. .e impact of knowing the disease and medication. One of the main objectives is to improve the greatly influences the roles played by patients at home, at monitoring of patients with the aim of also improving the work, and in their community [5]. quality of life of people. For this purpose, the HCI discipline .e articles [6, 7] agree that there are usability problems contributes to the design of usable interfaces that improve in health applications. .erefore, the participation of pa- the record of their evolution. tients in the design is very important, so that it can have an .e analysis of emotions has begun to acquire scientific impact on the state of patients through the use of the app. interest according to the possible applications, increasing the According to [6], patients are interested in using these new number of studies related to the emotional component and technologies, but current tools are not suitable for them. For its form of evaluation in different areas [9]. this reason, we aim at analyzing the data and give patients .e role of emotional communication in the context of control over the collection of their data and see how their HCI is very relevant and challenging when considering the different areas of technological application [10], like in the status affects the use of the application. In this work, the end users are cancer patients, being the example that focuses on remote medical care and assistive evaluation of all the possible influential factors on their technologies. In this area, there is an increasing interest in evolution (psychic, physical, treatments, and pharmacology) this type of research, which includes a synthesis of the effects fundamental. For a better design and development of the and emotion. Emotions affect human behavior and the application, we follow a patient-centered design, with pa- system, so when a user is using an application, their emo- tients and medical professionals in successive interviews tional state can affect the usability of the system [11]. with them. .e term mHealth is defined as the union of mobile .e objective of the present study is to perform an integral computing, medical sensors, and communication technol- follow-up of cancer patients through the use of a mobile ogies for health care [12]. MHealth is an upgrowing field application that allows registering the levels of physical and about the application of mobile technologies in health, which in recent years has emerged as an important segment psychological follow-up, as well as additional influential parameters in its evolution: treatments, medication, activities of telemedicine and its main objective is to improve health carried out, and quality of sleep among others. .is moni- services, integrating the benefits of mobility and ubiquity, toring allowed professionals to evaluate the effectiveness of typical of mobile systems, to the care treatments of tradi- the treatments, as well as to guide or personalize their se- tional health, pretending to bring health care to people and lection, identifying additionally the population at risk and not people to the health system. MHealth applications are thus improving the quality of life of patients. creating mechanisms for the exchange of information re- Additionally, parameters are recorded to take interaction lated to health care, even in remote and low-income areas, patterns into account, such as the choice changes made by due to the large area of coverage and social influence of the patient when replying to each question and the whole mobile telephone networks, becoming a factor strategic to reply time. .rough the study of the collected data, we made save lives [13]. In this context of mHealth, we will focus on an analysis of the relationship between user variables the follow-up of cancer patients. (concerning mood, sleeping quality, and pain) and the in- In addition to the intervention of the doctors in the teraction patterns. consultations with their patients, another relevant aspect is .e rest of the paper is structured as follows. .e related the optimization of their next consultations before doing work is studied in Section 2. Section 3 analyzes the materials them. In this way, they can improve the way in which the and methods that we use in this research. .e results of the medical professionals dictate a new treatment. .ere is an research are presented in Section 4. Section 5 discusses the application of medical treatment to optimize the consulta- results and the limitations. Finally, the conclusion and future tions of medical patients according to the known preferences work are mentioned in Section 6. and other selection criteria [14]. .at work refers to opti- mizing the appointments of the patients, the insurance payment options, the treatment in the medical facilities, and 2. Related Work other aspects of the interactions between patients. .e ap- .e Human-Computer Interaction (HCI) applied to the plication gathers data from the patient’s history and cal- culates a placement score in a medical center based on a field of health is increasingly essential in the development of interactive systems for medical professionals and patients plurality of parameters of the user’s medical history asso- ciated with the user’s medical history and schedules the [8]. .e work of medical professionals (doctors, psycholo- gists, etc.) can be simplified by the continuous use of medical care consultation at a medical center that provides technology to obtain information in real time from patients. the highest placement score in a medical center. Many more Patients can be supported by technology during their applications are being developed and can be downloaded treatment or recovery process, both in the diagnosis and through the Google Play Store for Android phone users. treatment phases, and in the case of having successfully Some of the mHealth systems are for some hereditary overcome the disease. Providing patients with these appli- diseases and disorders [15], cancer-based apps [16–19], or studies on the impact of stress or emotions on the interaction cations can be beneficial for them as they are more pleasant, or at least not as monotonous, in their treatment of the with the application [11, 20]. Journal of Healthcare Engineering 3 Android Mobile Informatics Application for Some contemplates the integral monitoring of patients. Patients Hereditary Diseases and Disorders (AMAHD) is a com- were able to see and manage both their medication and their plementary framework for medical practitioners and pa- appointments. For the part of the appointments that have tients. .e mobile application will help to sensitize and both with the psychologist and with the doctor, a section was complement the efforts of biomedical, medical, and bio- designed in which they can visualize the appointments in a informatics researchers working in the areas of inheritance calendar or in a detailed list, and they can be managed by research and genetics. AMAHD has proven to be a valuable them. .e part of the patient’s medication has also been resource for the research company in the battle against designed, in this way they can keep track of his prescriptions, hereditary diseases and disorders [15]. allowing them to visualize and manage them themselves; the medication can be viewed weekly or through a list with the .ere are some cancer-based applications. For instance, details. a smartphone-based pain management app for adolescents .e methodology used for the application Close2U with cancer provides patients pain management support was an agile methodology based on the iterative and based on their individual pain [16]. Another article evaluates incremental development of successive prototypes with the usability of their app and shows the results of the iterative increased functionality. .e application has evolved in its development of their app. Its authors inform other devel- phases of analysis, design, implementation, and evalua- opers and researchers in development, integration, and tion. .e successive refinements have allowed the evo- evaluation of mobile health apps and services that support lution of the prototypes and the increase of their cancer patients in managing their health-related issues [17]. functionality and quality. .e agile methodology may be .e goal of another app is to stabilize a daily functional the most appropriate for projects that suffer a high activity in breast cancer patients. App-using participants number of changes and need more control and com- could more frequently report adverse events, and those munication with the client in real time and allows both under supervision made fewer and more precise entries than adapting to problems that may arise and making the unsupervised participants [18]. .e last reviewed app is a necessary changes at the beginning of each phase, without smartphone app framework for segmented cancer care having to wait to finish all the actions. coordination, which provides both medical risk assessment Nowadays, it is crucial to incorporate the user into the and health care monitoring functions [17]. However, none of design and implementation cycle of a mobile application. In these apps considered information from interaction patterns addition, in the medical-scientific research community, to enrich the extracted information, as the current work there is a growing thought that states that the control and does. prevention of cancer must incorporate a communication Besides implemented applications, some papers have with the patient [10]. In this way, the benefit of current studied the impact of emotions over the medical applications medical discoveries in diagnosis and treatment is maxi- and how emotions can be related to the usability of the mized, particularly in the emerging era of personalized system that they are interacting with [11, 20]. Defining the medicine. Although patient-medical communication has stress level of the user can train the system in such a way that focused on results such as patient satisfaction, under- it could not only detect the user’s stress level but can also standing, and assessment, we must strengthen the under- modify itself accordingly, thereby increasing the usability of standing of how these can impact on their attitude and the the system and the user satisfaction [11]. In [20], it is shown results of the disease [10]. .roughout all phases of devel- that experiencing negative emotions during the use of the opment, we have worked in cooperation with the psy- system can negatively influence important user behaviors, chologists of the AECC (Spanish Association against including the client’s decisions regarding the application. Cancer) in Teruel, who were present throughout the cycle of In conclusion, monitoring and reminder applications design, implementation, and evaluation of the product. have been helpful for users in assisting them in monitoring To design and develop the application, a conceptual and keep track of the patient’s health care records as well as framework was first made; then the necessary algorithms their medication intake, and there are some studies that for data capture were designed and the software tools show how the emotions can affect the usability of mobile necessary to design it were selected. .e developed ap- applications. .is motivates the development of Close2U plication allows one to treat and use all the information application with the intention to help the medical practi- collected about physical condition, activities, treatments, tioners with the analysis that the app can bring them and and so on. Each element of the follow-up was funda- help them in improving the treatment for the patient. In mental to know the overall situation of the patient, addition, a novelty of Close2U app is that it shows that the maintaining their privacy. Information was collected and, analysis of interaction patterns can enrich self-reported user in order of priority, actions to be carried out were defined. information. .is work analyzes the correlations among cancer-patient variables and interaction-pattern variables. 3.1.1. User Interface of Close2U App. .is section presents the user interface showing the most relevant screens of the 3. Materials and Methods final implementation of Close2U app. .e changes that 3.1. Materials: Close2U App. .e main material for this occurred in the successive evaluations will be explained, research is the mobile application Close2U. .e application since changes were made throughout the implementation, 4 Journal of Healthcare Engineering (a) (b) Figure 1: Screens of the user interface 1: (a) mood selection and (b) sleep quality selection. with the help of the evaluation we made together with the Another valued aspect was the parts where the patient psychologists and the cancer patients. felt pain or discomfort, and a screen was made with which Figure 1(a) shows the screen in which the patient is asked the patient could visualize a body in which a pain could be indicated by buttons to make it more visual. .is is about their mood. It is a question that had many changes since at the beginning, the emotions were not taken into account, intended to be more intuitive for the patient as seen in Figure 2(c), since at first there was no image but simply only the mood, which limited the patient to specifying his mood. Once the emotions were introduced, it was possible to buttons. clarify in more detail how they feel, opening the way to a new “mood register” called “Undecided”; this state of mind de- 3.1.2. Internal Functioning of Close2U App. .e purpose of pends on the emotions that the patient selects. When the the design phase was to ensure that the developed appli- patient selects various moods, the medical professional can cation meets the requirements of the end user before the see the situation of patient’s indecision. prototype is translated into the production application. We In the screen of Figure 1(b), one can see the question used the Unified Modeling Language (UML) for defining the about sleep quality. .is screen was modified several times design diagrams. To carry out the plan of user activities, the psychol- according to the requirements and evaluation meetings established by the psychologists of the AECC Teruel; it was ogists of the AECC Teruel were taken into consideration. decided to show the patient in a numerical form from 0 to In this way, the interactions that the user must perform 10, since it is more comfortable and clearer in terms of with the application were identified. Figure 3 shows how usability. .is same change was also made with the level of patient interacts with the application when carrying out pain, which is presented later in Figure 2(b). the surveys. In each question, the app checks that the Figure 2(a) shows the options to assess whether you feel answer is valid and continues to the next question until pain or discomfort, or do not feel both. Initially only the pain the end of the questionnaire. .e replies are sent through was registered, but later considering the comments from the API, and it will be verified that the survey has been patients and psychologists, it was later decided to add a registered. In the user interface, the app confirms the differentiation between pain and discomfort. patient that the survey has been successfully registered. Journal of Healthcare Engineering 5 (a) (b) (c) Figure 2: Screens of the user interface 2: (a) pain/discomfort selection, (b) pain level selection, and (c) body parts with pain selection. with its emulator to perform the patient part. .e imple- For the selection of the questions to be shown in the survey, a previous analysis was carried out and they were mentation and validation were carried out over twelve refined through successive evaluations. Next, the interface months. associated with each of them was designed, as described in .e validation of the application has been made from the Section 3.1.1. For this stage, the flow of activities and their requirements analysis phases until the final implementation order were designed, so that in this way the patient will with the medical professionals of the AECC Teruel. .e first “connect” better with the survey and neither feels bad tests with patients allowed us to correct the existing con- when doing it nor gets bored along with it. ceptual and design errors. In the next stage of the design, once the activities were Once the first functional prototype of the application defined, a sequence diagram was defined, and Figure 4 presents was ready, tests were carried out with the users. Based on this diagram, in which we show how the surveys work internally them, we got back to the design phase and improved the design. .rough the evaluation carried out, the interfaces in the application. For all the questions shown on the screen, in parallel when the patient answers each question, the change is were modified to improve aspects of the usability and added to that answer. .en, when the patient confirms their functional application, for example, so that this would be answer, it goes to the next question. For the first and last intuitive for the patient. In addition, being able to assess the questions, there are special sequences. In the first question, the mood of the users during the use of the live system had a start time of the survey is recorded and in the last question the practical meaning for the design and improvement of our total time is calculated, the response is sent to the server, and application. finally, the survey ends. At the sequence diagram, the objects are shown as 3.2. Methods lifelines along the survey and with their interactions over time represented as arrows from the origin lifeline to the 3.2.1. Participants. .ere were 23 users that voluntarily participated in this study, without any economical com- destination lifeline. Sequence diagrams are appropriate for showing which objects communicate with each other and pensation. .ey were 50.21 years old in average with a the messages that trigger those communications. standard deviation (SD) of 10.09. .ey were 6 males and 17 For the implementation, the IDE (Integrated Devel- females. .e sample included 7 cancer patients and 16 opment Environment) Android Studio was used, together healthy people. 6 Journal of Healthcare Engineering Question Web service Yes Patient Is the last question? Show question No First question setInitTime() Show question? Parallel setAnswer () Select answers addAnswerChange () check (RegSurvey) Yes Valid answers? goNext (RegSurvey) No Last question surveyTime () Show error post (RegSurvey) send and receive API data Next question goFinish (RegSurvey) Patient Question Web service Figure 4: Sequence diagram of Close2U app. Send data to the API (i) Category 1: overwhelmed, agitated, ashamed, de- pressed, angry, exhausted, hurt, scared, alone, ap- No e survey has been registered? prehensive, defeated, desperate, despondent, exhausted, helpless, angry, frustrated, impatient, pessimistic, self-critical, irritated, defensive, de- Show survey done spised, and resentful (ii) Category 2: guilty, stressed, angry, bad, grumpy, dizzy, nervous, sad, discouraged, apathetic, worried, anxious, agitated, hurt, disappointed, distressed, Figure 3: Activity diagrams of Close2U app. nostalgic, offended, hungry, lacking confidence, impotent, repentant, insecure, and rejected 3.2.2. Measures. We evaluated sleep quality with the self- (iii) Category 3: bored, confused, asleep, busy, pensive, reported question “How did you sleep?” replied in a 0–10 tired, neutral, unequal, confused, hesitant, lazy, disconnected, reserved, indifferent, apprehensive, range (zero meaning the worst sleep quality and ten meaning the best). We determined the existence of pain conflicted, disconnected, reserved, indifferent, scattered, restless, sensitive, and vulnerable with the question “Have you noticed any pain or dis- comfort in the last 12 hours?” with an answer from “yes,” (iv) Category 4: well, happy, awake, relieved, cared, “no,” and “discomfort” options. If the user replied affir- empathetic, not critical, confident, calm, and sincere matively, then the app asked, “What level of pain/dis- (v) Category 5: super, proud, satisfied, confident, alive, comfort do you have now?” which was answered in a 1–10 enthusiastic, strong, encouraged, excited, grateful, range (one, minimum pain level and ten, maximum). .e hopeful, and open-minded app recorded zero level if the user had previously an- swered to feel neither pain nor discomfort. We evaluated We labeled the responses with more than one mood as the mood by asking the user “How are you feeling?” with a undecided. We assessed the indecision for each question multiple response among 97 moods classified in five mood with the number of times that the user changes their reply categories (1: horrible, 2: discouraged, 3: normal, 4: an- before submitting it. We applied this measure to (a) sleep imated, and 5: radiant), categorized by the medical pro- quality, (b) whether the user felt pain, (c) the level of this fessionals with who we worked in every step of the design pain, and (d) the mood. and development of the app. .ese moods were the fol- We used System Usability Scale (SUS) [21] to measure lowing ones: the usability of the app. Journal of Healthcare Engineering 7 Table 1: Relation of mood changes and time of realization. Average Undecided Radiant Animated Normal Discouraged Horrible Num. changes when selecting mood (#) 1.57 0.22 0.39 0.24 0.67 1.50 Survey completion time (mm:ss) 00:57 01:08 00:57 01:00 01:16 01:36 2.00 04:00 1.75 03:30 1.50 03:00 1.25 02:30 1.00 02:00 0.75 01:30 0.50 01:00 0.25 00:30 0.00 00:00 Undecided Radiant Animated Normal Discouraged Horrible Mood Time Changes Figure 5: Relation of mood changes and time of realization. 3.2.3. Protocol. Firstly, we followed a user-centered design. usually take more changes before selecting their final state. .ree psychologists familiar with cancer tested the Close2U Also, in this case, they select more than one state of mind. app during the whole development process and provided .e average of the number of changes in this case is 1.57. feedback. .e developer improved the app following their Regarding the total time of the survey, this was generally recommendations until they were satisfied, to achieve a high longer when the patient felt worse. level of usability. In this way, the functionalities of the app In a first approximation, it can be estimated that the were enriched according to the common specific needs of indecision corresponds to positive moods, since the average cancer patients and their doctors. completion time of the undecided survey was similar to the .e app was uploaded to Google Play, which is the main times of positive states. store of apps for Android. We encouraged cancer patients to Table 2 shows the average of the changes made by the use the app in a noncontrolled environment, advising them patient each time he had to choose the level of pain and the to use the app regularly. time in which he performs the survey. We measured the collected information about users What we can observe in the graph of Figure 6 is (related to mood, sleep, and pain) and the information from an increase in the number of changes that the patient made interaction patterns (number of changes when selecting when they had more pain, as in the time of conducting the replies to specific questions and the global survey time). We survey that is higher when the patient chose a higher level of analyzed the correlations among all these variables. pain. .is relation was also observed in other experiments, as We measured the usability of the app at the end of the one can see in the next presented graph. study with SUS scale. Table 3 shows an average of the changes made by the patient when he had to decide if he had pain or not, and in 4. Results the case of having pain if it is discomfort or pain and the time in which he performed the survey. Table 1 shows (a) the average of the changes made by the In the graph of Figure 7, it can be seen to what extent it patient each time he had to choose the emotions that mark affected that the patient did not have pain, and we found that his mood and (b) the time the survey took. it took less than half the time to complete the survey. When In Figure 5, it can be observed that the patient modifies comparing pain and discomfort, there were not such dif- his response in a greater number of occasions when his ferences in the time to carry out the survey. mood is negative. When the users are undecided, they Num. changes when selecting mood (#) Survey completion time (mm:ss) 8 Journal of Healthcare Engineering Table 2: Relation of pain level changes and time of realization. Average 1 2 3 4 5 6 7 8 9 10 Num. changes when selecting pain level (#) 0.08 0.42 0.78 0.42 0.93 0.67 0.31 0.61 1.00 1.43 Survey completion time (mm:ss) 01:40 01:29 01:19 01:30 01:16 01:17 00:59 01:43 03:35 02:30 2.00 04:00 1.75 03:30 1.50 03:00 1.25 02:30 1.00 02:00 0.75 01:30 0.50 01:00 0.25 00:30 0.00 00:00 123456789 10 Pain level Time Changes Figure 6: Relation of pain level changes and time of realization. Table 3: Relation of pain changes and time of realization. Average No pain Pain Discomfort Num. changes when selecting pain (#) 0.00 0.04 0.13 Survey completion time (mm:ss) 00:49 01:22 01:24 Table 4 shows an average of the changes that the patient not know whether the correlations would be direct or inverse makes when deciding the level of sleep and the time in which beforehand. the survey is performed, for each sleeping-quality level. Regarding the user variables and interaction patterns Figure 8 shows this information graphically. One can ob- concerning the numbers of changes, the test found significant serve that with the exception of a peak in level 1, the time of correlations in the pairs (a) mood and number of changes when selecting mood, (b) sleep and number of changes when conducting the survey barely changed. By contrast, the number of changes was higher when the patient had slept selecting sleep, (c) pain and number of changes when selecting better. pain, (d) pain and number of changes when selecting pain level, What we can observe in the graph of Figure 8 is that, with (e) pain level and number of changes when selecting mood, (f) the exception of the peak that exists in level 1 of sleep, the pain level and number of changes when selecting sleep, (g) pain time of accomplishment of the survey hardly changed. On level and number of changes when selecting pain, and (h) pain the other hand, the number of changes increased as the level and number of changes when selecting pain level. patient had slept better. Concerning the user variables among each other, the test found In order to determine if there were statistically signifi- significant correlations in all the possible pairs among the user cant correlations, we conducted the Pearson correlation test variables (1) mood, (2) sleep, (3) pain, and (4) pain level. All the between user variables (i.e., mood, sleeping quality, the significant correlations between mood-related variables and other variables had a negative correlation coefficient. However, existence of pain, and pain level) and the number of changes when selecting these variables. Table 5 shows the results of all the significant correlations among non-mood-related var- this test. We considered 497 cases (considering a case each iables had a positive correlation coefficient. time a user performed the complete survey in the app) for In order to determine if the correlations between this analysis. We considered 2-tailed significances as we did survey time and user variables were significant, we Num. changes when selecting pain level (#) Survey completion time (mm:ss) Journal of Healthcare Engineering 9 2.00 04:00 1.75 03:30 1.50 03:00 1.25 02:30 1.00 02:00 0.75 01:30 0.50 01:00 0.25 00:30 0.00 00:00 No pain Pain Discomfort Type of pain Time Changes Figure 7: Relation of pain changes and time of realization. Table 4: Relation of sleep quality changes and time of realization. Sleeping quality 0 1 2 3 4 5 6 7 8 9 10 Num. changes when selecting sleep quality (#) 0.00 0.00 0.14 0.27 0.15 0.23 0.12 0.56 0.75 0.58 1.07 Survey completion time (mm:ss) 01:06 04:25 01:19 00:51 00:53 01:05 00:54 01:07 01:11 00:57 01:05 2.00 05:00 1.75 04:00 1.50 1.25 03:00 1.00 02:00 0.75 0.50 01:00 0.25 0.00 00:00 0 1 23456789 10 Sleep quality Time Changes Figure 8: Relation of sleep quality changes and time of realization. conducted Pearson’s correlation test, and Table 6 shows Regarding usability, the average result of SUS test was the results. .is test found significant correlations be- 69.2 in the standard range of 0–100, and the standard tween (a) survey time and (b) the pain-related variables deviation (SD) was 20.0. Figure 9 shows the results of the including pain and pain level. individual items from the SUS scale. Num. changes when selecting sleep quality (#) Num. changes when selecting pain (#) Survey completion time (mm:ss) Survey completion time (mm:ss) 10 Journal of Healthcare Engineering Table 5: Pearson’s correlation test between user variables and number of changes when selecting the response for these variables, respectively. Num. Num. Num. Num. changes changes changes when changes Pain when selecting Mood Sleep when Pain selecting when level mood pain selecting mood selecting pain level sleep Pearson’s ∗∗ ∗∗ ∗∗ ∗∗ 1 −0.172 −0.133 −0.054 −0.158 −0.080 −0.335 −0.035 correlation Mood Sig. 0.000 0.003 0.234 0.000 0.074 0.000 0.438 (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗ −0.0172 1 0.084 0.166 −0.052 0.023 0.115 0.035 Num. changes correlation when selecting Sig. 0.000 0.061 0.000 0.244 0.608 0.010 0.433 mood (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗∗ ∗∗ −0.133 0.084 1 0.300 0.121 0.027 0.125 0.037 correlation Sleep Sig. 0.003 0.061 0.000 0.007 0.548 0.005 0.412 (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗∗ ∗∗ −0.054 0.166 0.300 1 0.053 −0.025 0.139 0.402 Num. changes correlation when selecting Sig. 0.234 0.000 0.000 0.234 0.581 0.002 0.000 sleep (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗∗ ∗∗ ∗∗ −0.158 −0.052 0.121 0.053 1 0.194 0.703 0.393 correlation Pain Sig. 0.000 0.244 0.007 0.234 0.000 0.000 0.000 (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ −0.080 0.023 0.027 −0.025 0.194 1 0.149 0.049 Num. changes correlation when selecting Sig. 0.074 0.608 0.548 0.581 0.000 0.001 0.274 pain (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗ ∗∗ ∗∗ ∗∗ ∗∗ ∗∗ −0.335 0.115 0.125 0.139 0.703 0.149 1 0.371 correlation Pain level Sig. 0.000 0.010 0.005 0.002 0.000 0.001 0.000 (2-tailed) N 497 497 497 497 497 497 497 497 Pearson’s ∗∗ ∗∗ ∗∗ Num. changes −0.035 0.035 0.037 0.402 0.393 0.049 0.371 1 correlation when selecting Sig. mood pain 0.438 0.433 0.412 0.000 0.000 0.274 0.000 (2-tailed) level N 497 497 497 497 497 497 497 497 ∗∗ ∗ Correlation is significant at the 0.01 level (2-tailed). Correlation is significant at the 0.05 level (2-tailed). Table 6: Pearson’s correlation test between survey time and user variables. Survey time Mood Sleep Pain Pain level ∗∗ ∗∗ Pearson’s correlation 1 −0.049 0.007 0.335 0.323 Survey time Sig. (2-tailed) 0.272 0.874 0.000 0.000 N 497 497 497 497 497 ∗∗ ∗∗ ∗∗ Pearson’s correlation −0.049 1 −0.133 −0.158 −0.335 Mood Sig. (2-tailed) 0.272 0.003 0.000 0.000 N 497 497 497 497 497 Journal of Healthcare Engineering 11 Table 6: Continued. Survey time Mood Sleep Pain Pain level ∗∗ ∗∗ ∗∗ Pearson’s correlation 0.007 −0.133 1 0.121 0.125 Sleep Sig. (2-tailed) 0.874 0.003 0.007 0.005 N 497 497 497 497 497 ∗∗ ∗∗ ∗∗ ∗∗ Pearson’s correlation 0.335 −0.158 0.121 1 0.703 Pain Sig. (2-tailed) 0.000 0.000 0.007 0.000 N 497 497 497 497 497 ∗∗ ∗∗ ∗∗ ∗∗ Pearson’s correlation 0.323 −0.335 0.125 0.703 1 Pain level Sig. (2-tailed) 0.000 0.000 0.005 0.000 N 497 497 497 497 497 ∗∗ Correlation is significant at the 0.01 level (2-tailed). 5 In the significant correlations, the sign of correlation co- efficients showed that mood-related variables were inversely correlated with sleep-related variables and pain-related vari- ables. .e inverse correlation between mood and sleep variables makes sense because the pain negatively influenced on the mood of users, as the pain probably did not let them sleep well. .e correlation between the survey time and the pain 2 (including the existence and its level) revealed that the in- teraction patterns also provided useful information about pain. .e survey of the app dedicated a large part to the pain, since the user was asked about the specific locations of the pain in a silhouette-based interface. Users spent more time depending on when they had a higher level of pain. .us, 123456789 10 future applications for cancer patients with silhouette-based Question numbers of SUS interfaces could indirectly help in estimating the pain level. Figure 9: Results of the SUS scale. One of the limitations of this research is the small sample size of participants. .is sample size only allowed us to detect correlations with large effect sizes. .us, we may have not detected correlations with small or medium effect sizes. 5. Discussion We detected a correlation between mood and the number of 6. Conclusion and Future Work changes when selecting the answers. .is reveals that the number of changes when selecting mood is relevant as it is A mobile application has been developed that performs a correlated with useful information. .is implies that cancer complete follow-up of the patient. .is article has focused on apps can collect enriched information from users by counting the self-reported moods, sleeping quality, and pain when the the number of changes when selecting a response in questions. patient is using our application, to provide a tool for psy- .erefore, interaction patterns provided relevant information chologists to (a) improve their treatment, (b) improve ap- concerning moods. pointments with them, (c) carry out a daily monitoring of .e results revealed that users were doubtful (i.e., they patients, and (c) be able to communicate with them outside changed the replies more times) when selecting (a) negative face-to-face consultations. .is increased the patients’ sense moods, (b) high-quality levels of sleep, and (c) high levels of of being taken care of daily, since they did not feel alone. pain. Notice that we distinguished between low and high .is article shows (1) how a mobile health application values for each user variable, and this was concluded from can assist patients in being more active in managing their the sign of correlation coefficients. For example, in the case care, (2) how the emotions affect when completing the of moods, this may reveal that users perceived negative survey, and (3) taking more time or making more changes emotions as a combination of some basic emotions, in ac- usually depend on their mood, pain, or sleep quality at that cordance to the theory of Ekman [22] about basic emotions, time. Our research revealed that interaction-patterns vari- rather than for positive emotions. ables provided useful information from some user variables, Another relevant finding was that pain level was sig- proved with the statistically significant correlations detected nificantly correlated with all the other studied variables in this study. including the user variables and the numbers of changes .is work motivates future research concerning the when selecting any user variable. Hence, pain level was the extraction of implicit information from interaction pat- variable with most relevant information due to its corre- terns in cancer monitoring apps. .e correlation between lations with all the other variables. user variables and interaction patterns triggers a Response in 5-point Likert scale 12 Journal of Healthcare Engineering [2] S. S. MS Mohammed, J. Lloret, and I. Bosch, “Systems and promising research line about getting patients’ informa- WBANs for controlling obesity,” Journal of Healthcare En- tion from their patterns when interacting with apps, being gineering, vol. 2018, Article ID 1564748, 21 pages, 2018. able to enrich the collected information with this non- [3] J. Lloret, L. Parra, M. Taha, and J. Tomas, ´ “An architecture and self-reported information. .is research could also lead to protocol for smart continuous eHealth monitoring using 5G,” take implicit-extracted mood to adapt the interface to the Computer Networks, vol. 129, pp. 340–351, 2017. needs of patients. [4] S. Loibl and B. Lederer, “.e importance of supportive care in According to medical professionals, the app helped them to breast cancer patients,” Breast Care, vol. 9, no. 4, pp. 230-231, collect relevant data about the disease more easily than if they had had to collect these manually. In the future, we plan to [5] F. Cardoso, N. Bese, S. R. Distelhorst et al., “Supportive care analyze how treatment improves and how waiting times are during treatment for breast cancer: resource allocations in reduced before face-to-face consultations. low- and middle-income countries. A Breast Health Global In the future, it is intended to perform the noninvasive Initiative 2013 consensus statement,” 9e Breast, vol. 22, no. 5, pp. 593–605, 2013. recording and measurement of the evolution of patients, by [6] U. Sarkar, G. I. Gourley, C. R. Lyles et al., “Usability of means of devices or sensors, to improve both their treatment commercially available mobile applications for diverse pa- and their quality of life. For this purpose, we will search tients,” Journal of General Internal Medicine, vol. 31, no. 12, available devices or sensors that will support the capture of pp. 1417–1426, 2016. evaluable parameters that affect patients, at the same time [7] H. Fu, S. K. McMahon, C. R. Gross, T. J. Adam, and that our application will collect data. Initially, smart sensors J. F. Wyman, “Usability and clinical efficacy of diabetes mobile and wristbands will be used to record the heart rate and the applications for adults with type 2 diabetes: a systematic Electrocardiograms (ECG), blood pressure, and Heart Rate review,” Diabetes Research and Clinical Practice, vol. 131, Variability (HRV). We will also explore the registry of other pp. 70–81, 2017. parameters such as oxygen level in the blood, skin and body [8] D. DAuria, F. Persia, and B. Siciliano, “Human-computer temperature, and breathing patterns among others. interaction in healthcare: how to support patients during their wrist rehabilitation,” in Proceedings of the 2016 IEEE Tenth International Conference on Semantic Computing (ICSC), Data Availability pp. 325–328, IEEE, Laguna Hills, CA, USA, 2016 February. [9] G. Sacca Espinoza, “Estrategia para el reconocimiento de .e data are not available to avoid privacy infringement. emociones colectivas en entornos sociales basados en tic strategy for recognizing collective emotions in social environments based on ict,” M.S. thesis, Universidad Autonoma ´ de Madrid, Conflicts of Interest Madrid, Spain, 2015. [10] A. Esposito, A. M. Esposito, and C. Vogel, “Needs and .e authors declare that there are no conflicts of interest challenges in human computer interaction for processing regarding the publication of this paper. social emotional information,” Pattern Recognition Letters, vol. 66, pp. 41–51, 2015. [11] S. Saxena, P. Saxena, and S. K. Dubey, “Various levels of Acknowledgments human stress & their impact on human computer interac- tion,” in Proceedings of the 2013 International Conference on .e authors would like to thank the psychologists of the Human Computer Interactions (ICHCI), pp. 1–6, IEEE, AECC of Teruel for their willingness to help them in im- Chennai, India, August 2013. proving the application. .is work was funded by the re- [12] R. S. H. Istepanian, E. Jovanov, and Y. T. Zhang, “Guest search projects “New Technologies in Cancer Monitoring editorial introduction to the special section on m-health: and Treatment” and “Advances in Research, Diagnosis and beyond seamless mobility and global wireless health-care Monitoring of Bone Cancer Patients as a Primary or Sec- connectivity,” IEEE Transactions on Information Technology ondary Tumor” by the University Foundation “Antonio in Biomedicine, vol. 8, no. 4, pp. 405–414, 2004. Gargallo.” .is work was partly funded by the Spanish [13] V. W. Consulting, mHealth for Development: 9e Opportunity Science, Innovation and University Ministry (MCIU), the of Mobile Technology for Healthcare in the Developing World, UN Foundation-Vodafone Foundation Partnership, Wash- National Research Agency (AEI), and the EU (FEDER) ington DC, USA, 2009. through the contract RTI2018-096986-B-C31 and by the [14] D. R. Albro, K. A. Darby, R. J. Dudzinski et al., “Medical Aragonese Government and the EU through the FEDER treatment application for optimizing medical patient visits 2014–2020 “Construyendo Europa desde Aragon” ´ action based on known preferences and other selection criteria,” U.S. (Group T25_17D). .ey also acknowledge the partial sup- Patent Application No. 10262384, JUSTIA Patent, Mountain port from CYTED (ref. 518RT0558) and MCIU (TIN2017- View, CA, USA, 2019. 88327-R). [15] O. Oluwagbemi, F. Oluwagbemi, and C. Ughamadu, “An- droid mobile Informatics application for some hereditary diseases and disorders (AMAHD): a complementary frame- References work for medical practitioners and patients,” Informatics in Medicine Unlocked, vol. 2, pp. 38–69, 2016. [1] L. Garc´ıa, J. Tomas, ´ L. Parra, and J. Lloret, “An m-health application for cerebral stroke detection and monitoring [16] L. A. Jibb, B. J. Stevens, P. C. Nathan, E. Seto, J. A. Cafazzo, and J. N. Stinson, “A smartphone-based pain management using cloud services,” International Journal of Information Management, vol. 45, pp. 319–327, 2019. app for adolescents with cancer: establishing system Journal of Healthcare Engineering 13 requirements and a pain care algorithm based on literature review, interviews, and consensus,” JMIR Research Protocols, vol. 3, no. 1, p. e15, 2014. [17] J. Mirkovic, D. R. Kaufman, and C. M. Ruland, “Supporting cancer patients in illness management: usability evaluation of a mobile app,” JMIR mHealth and uHealth, vol. 2, no. 3, p. e33, [18] M. Egbring, E. Far, M. Roos et al., “A mobile app to stabilize daily functional activity of breast cancer patients in collab- oration with the physician: a randomized controlled clinical trial,” Journal of Medical Internet Research, vol. 18, no. 9, p. e238, 2016. [19] T. He, R. Ogunti, M. Puppala et al., “A smartphone app framework for segmented cancer care coordination,” in Proceedings of the 2016 IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI), pp. 372–375, IEEE, Las Vegas, NV, USA, February 2016. [20] M. Hibbeln, T. Martin, J. L. Jenkins et al., “How is your user feeling? Inferring emotion through human-computer inter- action devices,” MIS Quarterly, vol. 41, no. 1, pp. 1–21, 2017. [21] J. Brooke, “SUS-A quick and dirty usability scale,” Usability Evaluation in Industry, vol. 189, no. 194, pp. 4–7, 1996. [22] P. Ekman, “An argument for basic emotions,” Cognition & Emotion, vol. 6, no. 3-4, pp. 169–200, 1992.

Journal

Journal of Healthcare EngineeringHindawi Publishing Corporation

Published: Jul 15, 2020

There are no references for this article.