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

Learn More →

Associations of thigh muscle fat infiltration with isometric strength measurements based on chemical shift encoding-based water-fat magnetic resonance imaging

Associations of thigh muscle fat infiltration with isometric strength measurements based on... Background: Assessment of the thigh muscle fat composition using magnetic resonance imaging (MRI) can provide surrogate markers in subjects suffering from various musculoskeletal disorders including knee osteoarthritis or neuromuscular diseases. However, little is known about the relationship with muscle strength. Therefore, we investigated the associations of thigh muscle fat with isometric strength measurements. Methods: Twenty healthy subjects (10 females; median age 27 years, range 22–41 years) underwent chemical shift encoding-based water-fat MRI, followed by bilateral extraction of the proton density fat fraction (PDFF) and calculation of relative cross-sectional area (relCSA) of quadriceps and ischiocrural muscles. Relative maximum voluntary isometric contraction (relMVIC) in knee extension and flexion was measured with a rotational dynamometer. Correlations between PDFF, relCSA, and relMVIC were evaluated, and multivariate regression was applied to identify significant predictors of muscle strength. Results: Significant correlations between the PDFF and relMVIC were observed for quadriceps and ischiocrural muscles bilaterally (p = 0.001 to 0.049). PDFF, but not relCSA, was a statistically significant (p = 0.001 to 0.049) predictor of relMVIC in multivariate regression models, except for left-sided relMVIC in extension. In this case, PDFF (p = 0.005) and relCSA (p = 0.015) of quadriceps muscles significantly contributed to the statistical model with R = 0.548. adj Conclusion: Chemical shift encoding-based water-fat MRI could detect changes in muscle composition by quantifying muscular fat that correlates well with both extensor and flexor relMVIC of the thigh. Our results help to initiate early, individualised treatments to maintain or improve muscle function in subjects who do not or not yet show pathological fatty muscle infiltration. Keywords: Healthy volunteers, Magnetic resonance imaging, Muscle contraction (isometric), Muscle strength, Thigh * Correspondence: stephanie.inhuber@tum.de Department of Sport and Health Sciences, Technische Universität München, Georg-Brauchle-Ring 60/62, 80992 Munich, Germany Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Inhuber et al. European Radiology Experimental (2019) 3:45 Page 2 of 10 Key points CSA is not clear. To confirm a close relationship between Magnetic resonance imaging detects changes in thigh muscle PDFF and strength is clinically important. muscle composition by quantifying muscular fat. Thethigh musclesareoneofthe largestmusclegroupsof Muscular fat correlates well with extensor and flexor the human body and thus important target muscles to diag- strength at the thigh. nose and monitor different diseases affecting local and Muscular fat, not cross-sectional area, can predict whole-body muscle pathologies such as knee osteoarthritis, muscle strength in thigh muscles. NMD, sarcopenia, and tumour cachexia. Detecting and un- The interaction between muscular fat and strength derstanding changes in thigh muscle quality, which are could become the basis for a biomarker for muscle based on muscle fat compositions that correlate with quality and function. muscle strength, could support initiating early and indivi- dualised treatment protocols to maintain or improve Background muscle function prior to clinical diagnosis. Therefore, we Previous research has demonstrated alterations in fat investigated the association of thigh muscle fat with isomet- composition of thigh muscles due to various pathological ric strength measurement using chemical shift encoding- conditions primarily when they are already in a chronic based water-fat MRI and a rotational dynamometer in stage, including musculoskeletal disorders, metabolic dis- healthy volunteers. The hypothesis is that thigh muscle eases, and neuromuscular diseases (NMD) [1–9]. Further- PDFF improves the prediction of muscle strength measure- more, physical exercise and training have shown to entail ments beyond muscle CSA. measurable changes in the thigh musculature [10–12]. Such alterations can nowadays be captured non-invasively Methods by means of imaging, with magnetic resonance imaging Subjects (MRI) being at the forefront particularly thanks to the Twenty healthy volunteers (10 females; median age 27 ability to perform qualitative and quantitative assessments years, range 22–41 years) were recruited. Age between 20 of the human body fat composition in vivo [13]. and 45 years and body mass index (BMI) between 23 and For the purpose of assessing muscle fat composition, 33 kg/m were defined as inclusion criteria to reflect the anatomical T1- and T2-weighted MRI is conventionally average population with a rather broad BMI range (median applied [13]. Based on anatomical imaging, the cross- BMI of the study population 26.7 kg/m ,range 22.2–31.8 sectional area (CSA) of muscles can be calculated, which kg/m ). Completing the international physical activity ques- can be used as a structural measure of muscle hyper- tionnaire in short-form ensured that all subjects had a trophy or atrophy [14–16]. More advanced MRI-based moderate level of physical activity (referring to the scoring methods including magnetic resonance spectroscopy protocol: 600–1500 metabolic equivalent of task-min/week) and chemical shift encoding-based water-fat imaging en- [26, 27]. Exclusion criteria were (1) any history of high- able the extraction of parameters like the proton density performance sports, (2) any history of metabolic diseases, fat fraction (PDFF) [13, 17]. In contrast to conventional NMD, previous knee or thigh muscle injuries, (3) general T1- and T2-weighted MRI, chemical shift encoding- contraindications for MRI (e.g.,cochlearimplants),and (4) based water-fat MRI allows for robust quantitative and, implanted foreign bodies at the level of the upper leg. thus, more objective evaluation of muscle composition MRI and strength measurements were scheduled whilst simultaneously enabling anatomical assessment within 5 days to avoid any mismatch between the mea- [18–20]. In this context, thigh muscles represent a re- sured strength and fat parameters as obtained by MRI- gion of high interest for MRI investigation thanks to related analyses. In case that both parts of the measure- good magnetic field homogeneity, minimum motion ar- ment took place on the same day, MRI was carried out tefacts, and the knowledge about disease-characteristic before strength measurements with the intent to not features of some pathologies in these muscles. Addition- capture potential modifications in muscular structures ally, it is possible to perform precise strength measure- due to strength measurement. ments at thigh muscles, thus allowing for direct links to The study was approved by the local institutional review changes in quantitative MRI to provide insights into board and conducted in accordance with the Declaration muscle quality and (dys-)function [21]. of Helsinki. All subjects gave written informed consent on However, evidence of association between thigh MRI examinations, isometric strength measurements, and muscle fat composition and muscle strength is scarce publication of identifying images prior to participation in [22–25]. Inverse relationships were reported for the the study. Data are accessible upon request. PDFF and strength at the thigh in patients with NMD [1, 8]. However, whether the PDFF of extensor and MRI flexor thigh muscles improves the prediction of thigh Subjects underwent 3-T MRI of the bilateral thigh mus- isometric strength in extension and flexion beyond the cles (Ingenia, Philips Healthcare, Best, The Netherlands). Inhuber et al. European Radiology Experimental (2019) 3:45 Page 3 of 10 Scanning was performed in supine position using a multi-echo mDIXON fat quantification routine of the built-in-the-table 12-channel posterior coil and a 16- vendor. The PDFF maps were computed as the ratio of channel anterior coil. the fat signal over the sum of fat and water signals. The An axially-prescribed, six-echo three-dimensional spoiled axial PDFF maps of both stacks of each subject were gradient-echo sequence was acquired for chemical shift stored for segmentation. encoding-based water-fat separation with the following pa- rameters: repetition time (TR)/echo time (TE) min/ΔTE = Segmentation 6.4/1.1/0.8 ms; field of view 220 × 401 × 252 mm ;acquisi- Segmentations of the quadriceps femoris muscles and tion matrix 68 × 150, voxel size 3.2 × 2.0 × 4.0 mm ;fre- ischiocrural muscles of both sides were performed quency encoding direction left-to-right L/R, no sensitivity within the ten most central slices depicting the thigh encoding. Scan time was 1 min and 25 s per stack. The six muscles (Fig. 1). Segmentations were done on the PDFF echoes were acquired in a single TR using non-flyback (bi- maps using MITK (http://mitk.org/wiki/The_Medical_ polar) read-out gradients with two axial stacks being ob- Imaging_Interaction_Toolkit_(MITK); German Cancer tained consecutively to cover the entire thigh region from Research Center, Division of Medical and Biological In- the hip down to the superior edge of the patella. A flip formatics, Medical Imaging Interaction Toolkit, Heidel- angle of 3° was used to minimise T1 bias effects [18, 28]. berg, Germany). The thigh muscles were manually segmented on the first, fifth, and tenth slice. Polygonal Imaging-based fat quantification regions of interest (ROIs) were carefully placed in these The imaging data were first processed by applying a axial slices. We then used the two-dimensional phase error correction and a complex-based water-fat interpolation tool of MITK to obtain segmentations of decomposition considering a pre-calibrated seven-peak the remaining axial slices. These segmentations were fat spectrum and a single T2* [29, 30]. We used the manually corrected. The ROIs were placed at the outer Fig. 1 Chemical shift encoding-based water-fat magnetic resonance imaging (MRI) and placement of regions of interest (ROIs). a Representative proton density fat fraction (PDFF) map. b PDFF map with superimposition of manually segmented muscle compartments defined as ROIs: (1) right quadriceps muscle, (2) left quadriceps muscle, (3) right ischiocrural muscles, and (4) left ischiocrural muscles. The red lines around the thigh represent the segmentation of the entire thigh contour Inhuber et al. European Radiology Experimental (2019) 3:45 Page 4 of 10 muscle contour whilst carefully avoiding the inclusion of subcutaneous fat or muscle fat interfaces [2, 22]. Subse- quent to ROI placements, the CSA (in mm ) and PDFF (in %) of the quadriceps and ischiocrural muscles were ex- tracted separately for both sides by averaging the respect- ive values obtained from the ten consecutive slices per side. Additionally, the entire thigh contour was segmented in the same ten slices (Fig. 1), followed by calculation of the mean thigh CSA per subject. A relative CSA (relCSA, in %) of quadriceps and ischiocrural muscles was deter- mined by dividing the respective muscle CSA by the entire thigh CSA. All segmentations were performed by a radi- ologist with 8 years of experience in musculoskeletal im- aging. Good reproducibility of measurements following this approach was previously reported [22]. Isometric muscle strength measurements The maximum voluntary isometric contraction (MVIC) in single-joint knee extension and flexion was measured at the thigh bilaterally with a rotational dynamometer (IsoMed 2000, D&R Ferstl GmbH, Hemau, Germany). Substantiated by measuring the isometric peak torque in Newton per Fig. 2 Setup for measurements of the maximum voluntary isometric metre (in N*m ), the MVIC was produced in knee exten- contraction (MVIC) with a rotational dynamometer sion at 60° and knee flexion at 35°, which are the joint an- gles with the ideal muscular strength-length relationship to anticipate the real MVIC [31–34]. At the beginning of the transfer the measured values of maximum isometric measuring visit, the isokinetic rotational dynamometer was torque to the software (proEMG, Prophysics AG, calibrated exactly on subjects’ individual body dimension. Switzerland). The value the tested leg produces whilst Subjects were seated in an upright sitting position and were sitting in a resting position caused by the gravity was fixed by hip and shoulder belts as well as shoulder pads, measured in each testing step. The measured absolute and two adjustable straps were used to fix the leg at the extension and flexion MVICs (in N*m ) were adjusted for pad of the lever arm in the correct measuring position the individual BMI to obtain a relative MVIC (relMVIC, (Fig. 2). in N*m /kg) for the left and right thigh muscles. Following the individual calibrating, the subjects had to perform a standardised warm-up exercise. After 10 Reproducibility of muscle strength measurements min on a cycling ergometer at 70–80 rpm (revolutions All participants performed at least one initial training per minute) to activate the cardiovascular system, the session. The visit’s aim was to become familiar with the volunteers were seated in the rotational dynamometer measuring procedure and to train to afford and activate where ten dynamic repetitions with moderate intensity the maximum isometric strength. To ensure the repro- were performed. In the fixed measuring position, the ducibility of strength measurements, each subject per- subjects’ task was to extend or to flex the knee against formed at least five to eight repetitions with a MVIC the measuring pad on the back or front side of the lower and 3 min of rest in between. A verifiable increase and leg with the individual maximum contraction of quadri- decrease of the measured values was assured as the de- ceps or ischiocrural muscles. MVIC of each direction of cisive factor for a potentially needed second training visit movement (extension/flexion) was collected three times or the testing appointment. There were at least 3 days of with 3 min of recovery in between, and the best value of rest between the visits and subjects were instructed to muscle flexion and extension maximum isometric torque come to the measurements totally recovered (no physical (in N*m ) was respectively taken for data analysis [2, activity the 2 days before visits). 22]. The choice of the starting leg and the direction of movement were randomised. Statistical analyses SPSS (version 20.0; IBM SPSS Statistics for Windows, Strength data collection method Armonk, NY, USA) was used for all statistical analyses The force transducer is located in the inside of the rota- and generation of graphs. The level of statistical signifi- tional lever arm of the used rotational dynamometer to cance was set at p < 0.05 (two-sided). The Kolmogorov- Inhuber et al. European Radiology Experimental (2019) 3:45 Page 5 of 10 Smirnov test indicated normal distribution of BMI, Muscle strength measurements PDFF, relCSA, and relMVIC values, but not of age. RelMVIC in extension was higher in males compared to First, mean ± standard deviation was calculated for PDFF, females in the left (8.0 ± 1.2 N*m /kg versus 5.9 ± 1.0 relCSA, and relMVIC, separately for male and female sub- N*m /kg, p < 0.001) and right quadriceps muscles 3 3 jects as well as for the right and left thigh muscles. Compar- (8.8 ± 1.3 N*m /kg versus 6.4 ± 0.9 N*m /kg, p < 0.001; isons of these measures between genders were performed Table 1). Similarly, males had greater relMVIC in flexion by using unpaired t-tests. Furthermore, Pearson correlation than females in the left (4.0 ± 0.6 N*m /kg versus 3.0 ± coefficients (r) were calculated between PDFF, relCSA, and 0.5 N*m /kg, p < 0.001) and right ischiocrural muscles 3 3 relMVIC. Using these parameters, multivariate regression (4.3 ± 0.6 N*m /kg versus 2.7 ± 0.7 N*m /kg, p < 0.001; models were calculated to identify significant predictors of Table 1). Quadriceps muscles showed approximately thigh muscle strength. Parameters were included in the re- twice as much relMVIC than ischiocrural muscles at gression models for p < 0.05. both thighs in males (mean quadriceps 8.4 ± 1.2 N*m /kg versus mean ischiocrural 4.1 ± 0.6 N*m /kg) and females (mean quadriceps 6.2 ± 1.0 N*m /kg versus mean ischio- Results crural 2.9 ± 0.6 N*m /kg; Table 1). Proton density fat fraction and cross-sectional area measurements Mean muscle PDFF was lower in males than females in the Correlations and multivariate regression models left and right quadriceps muscles (2.7 ± 1.3% versus 3.6 ± Significant correlations were revealed between the relMVIC 1.3%, p = 0.162; 1.7 ± 1.3% versus 2.7 ± 1.3%, p = 0.105) and in extension and the PDFF of the left (r = -0.649, p =0.002) ischiocrural muscles (3.2 ± 1.6% versus 4.6 ± 2.0%, p =0.091; and right quadriceps muscle (r = -0.612, p = 0.004; Table 2, 2.6 ± 1.9% versus 4.9 ± 2.4%, p = 0.025; Table 1). There were Fig. 3). RelMVIC in flexion was correlated significantly with no significant differences in age or BMI between males and the PDFF of the left (r = -0.446, p = 0.049) and right ischio- females (p = 0.280 and 0.684, respectively). crural muscles (r = -0.676, p =0.001; Table 2,Fig. 3). Males showed greater relCSA than females in the quad- Furthermore, the relMVIC in extension was signifi- riceps muscles (26.5 ± 4.0% versus 21.3 ± 4.1%, p =0.010; cantly associated with the relCSA of the left (r = 0.585, 25.8 ± 4.2% versus 21.8 ± 3.7%, p =0.035) and ischiocrural p = 0.007) and right quadriceps muscle (r = 0.448, p = muscles at both sides (8.3 ± 2.4% versus 6.0 ± 2.4%, p = 0.048; Table 2, Fig. 4). There was no significant correl- 0.045; 7.5 ± 3.0% versus 6.6 ± 1.6%, p =0.418; Table 1). ation found between the relMVIC in flexion and relCSA Quadriceps muscles had a bigger relCSA than ischiocrural of the ischiocrural muscles for the left and right sides muscles for males and females (Table 1). (Table 2, Fig. 4). PDFF, but not relCSA, was a statistically significant (p = 0.001 to 0.049) predictor of relMVIC in multivariate Table 1 Proton density fat fraction (PDFF), relative cross- regression models, except for left-sided relMVIC in ex- sectional area (relCSA), and relative maximum voluntary tension. In this case, PDFF (p = 0.005) and relCSA (p = isometric contraction (relMVIC) in extension and flexion 0.015) of the quadriceps muscles contributed signifi- Males Females p value cantly to the statistical model with R = 0.548. adj PDFF and relCSA PDFF quadriceps (%) Left 2.7 ± 1.3 3.6 ± 1.3 0.162 Right 1.7 ± 1.3 2.7 ± 1.3 0.105 Table 2 Correlation between the proton density fat fraction PDFF ischiocrural (%) Left 3.2 ± 1.6 4.6 ± 2.0 0.091 (PDFF) or relative cross-sectional area (relCSA) and relative Right 2.6 ± 1.9 4.9 ± 2.4 0.025 maximum voluntary isometric contraction (relMVIC) in extension relCSA quadriceps (%) Left 26.5 ± 4.0 21.3 ± 4.1 0.010 and flexion (males and females together) Right 25.8 ± 4.2 21.8 ± 3.7 0.035 relMVIC relCSA relCSA ischiocrural (%) Left 8.3 ± 2.4 6.0 ± 2.4 0.045 Left Right Left Right Right 7.5 ± 3.0 6.6 ± 1.6 0.418 Extension PDFF r -0.649 -0.612 −0.284 − 0.308 p value 0.002 0.004 0.225 0.186 RelMVIC in extension and flexion relCSA r 0.585 0.448 –– relMVIC in extension (N*m /kg) Left 8.0 ± 1.2 5.9 ± 1.0 < 0.001 p value 0.007 0.048 Right 8.8 ± 1.3 6.4 ± 0.9 < 0.001 Flexion PDFF r -0.446 -0.676 -0.125 -0.173 p value 0.049 0.001 0.599 0.465 relMVIC in flexion (N*m /kg) Left 4.0 ± 0.6 3.0 ± 0.5 < 0.001 relCSA r 0.238 0.207 –– Right 4.3 ± 0.6 2.7 ± 0.7 < 0.001 p value 0.312 0.380 Data are expressed as mean ± standard deviation Inhuber et al. European Radiology Experimental (2019) 3:45 Page 6 of 10 Fig. 3 Correlation between the proton density fat fraction (PDFF) and relative maximum voluntary isometric contraction (relMVIC). Plots showing the association between the left or right relMVIC in extension or flexion (in N*m /kg) and the PDFF (in %) of the left or right quadriceps or ischiocrural muscles Discussion muscles as part of the MyoSegmenTUM_thigh database, This study used chemical shift encoding-based water-fat with average PDFF values of 3.71% and 3.93% for the right MRI at the thigh in healthy volunteers to extract the and left quadriceps muscle as well as 4.38% and 4.44% for PDFF and relCSA of quadriceps and ischiocrural mus- the right and left ischiocrural muscles by also focusing on cles bilaterally and a rotational dynamometer to assess healthy males and females. However, a distinction be- relMVIC in extension and flexion. Muscle PDFF was a tween genders has not been achieved in their study be- better predictor of the relMVIC than the relCSA. We cause only three healthy females were included in total observed significant differences in the relCSA of ischio- [2]. Thus, the PDFF values obtained in the present study crural and quadriceps muscles, but also in the PDFF of seem to be principally in the range of previously reported ischiocrural muscles between males and females. values. Grimm et al. [3] recently reported averaged PDFF values Alterations in fat composition of thigh muscles have between 5.6 and 6.9% in healthy young males; however, in been shown in the context of diseases such as musculo- their investigation, PDFF measurements were derived skeletal disorders, metabolic diseases, and NMD [1–9]. from all thigh muscles and without distinct exclusion of Specifically, patients suffering from different types of intermuscular tissue. Specifically for the quadriceps mus- muscular dystrophy, Pompe disease, sarcopenia, osteo- cles, a previous study [22] reported a mean intramuscular arthritis, or type 2 diabetes mellitus showed increased PDFF of 4.02% in a cohort of healthy males. Schlaeger PDFF or intramuscular fat when compared to healthy et al. [2] performed the segmentation of individual thigh controls [1–9]. Although it has been hypothesised that Inhuber et al. European Radiology Experimental (2019) 3:45 Page 7 of 10 Fig. 4 Correlation between the relative cross-sectional area (relCSA) and relative maximum voluntary isometric contraction (relMVIC). Plots showing the association between the left or right relMVIC in extension or flexion (in N*m /kg) and the relative CSA (in %) of the left or right quadriceps or ischiocrural muscles increased fatty infiltration or transformation of thigh knee extension and indicated that the quadriceps inter- muscles is related to decrease in muscle strength [21, muscular adipose tissue fraction and intramuscular 23], studies correlating parameters such as the PDFF or PDFF correlate significantly with physical strength, rep- absolute or relative CSA of thigh muscles with objective resented by the MVIC. measurements of muscle strength are rare. An inverse The present study confirms this first evidence of an relationship was revealed between the PDFF and interaction between muscular fat and strength by examin- strength at the thigh in patients with NMD [1, 8]. In ing the associations between the PDFF and relMVIC at these studies, the obtained strength measurements were the thigh in healthy subjects and by showing not only sig- acquired with a handheld myometer [1, 8]; however, the nificant correlations for the PDFF of the quadriceps mus- usage of more objective devices to test muscle strength, cles but also for ischiocrural muscles. Moreover, we were such as isokinetic dynamometers that enable robust as- able to show that, in contrast to the relCSA, the PDFF of sessments of the muscle strength of specific functional the quadriceps and ischiocrural muscles were significantly muscle groups, is still largely missing to evaluate the as- associated bilaterally with relMVIC in knee extension and sociation between fat components and thigh muscle flexion. function. One previous study [22] focusing on the fat- These new insights in the interactions of fat and strength interaction of quadriceps muscles by only test- strength parameters in human muscles demonstrate the ing a small cohort of healthy subjects used an isokinetic importance of muscle quality, consisting of the individual dynamometer for isometric strength measurement in number of muscular contractile elements, the specific fatty Inhuber et al. European Radiology Experimental (2019) 3:45 Page 8 of 10 infiltration as well as the strength capacity [21]. The concerning muscle function and its role as an indicator muscle quality might be able to predict and flesh out to analyse and define muscle quality and fat-induced loss muscle (dys-)function a lot better than the relCSA does. of muscle function as well as to develop corresponding Of note, there is no significant correlation between the training programmes dealing with these ratios. As a per- relMVIC in flexion and the relCSA of the ischiocrural spective, the role of physiological CSA, muscular penna- muscles for the left and right side. The anatomical differ- tion angle, and muscle volume could complement and ence (quadriceps muscles: one-joint muscles; ischiocrural specify muscle quality [32, 35]. muscles: multi-joint muscles) as well as the variations con- Concerning methodology, chemical shift encoding- cerning the muscular structure and activation should be based water-fat MRI is confirmed as a fast method considered when interpreting this finding [35]. To specify which can be added to routine MRI protocols of the the knowledge of muscle quality and (dys-)function in the thigh region to visualise and quantify muscle quality and future, further studies are needed to include and prove the possibly forecast deficits in muscle strength and func- meaning of the muscle volume and in particular the tion, leading over to subsequent specific training pro- physiological CSA, the probably most powerful predictors grammes. In this context, PDFF derived from chemical concerning muscular strength [36–38]. shift encoding-based water-fat MRI is a more robust and Our observations seem to be in accordance with the objective approach compared to the semi-quantitative, finding of paraspinal PDFF being significantly correlated post-acquisitional analysis of conventional images [13]. with the relMVIC in extension and flexion of the trunk, However, the method requires compensation for con- whereas paraspinal mean CSA only showed significant founding factors, reflected by T2* decay, a potential correlations with relMVIC in flexion in a previous study quantification bias due to multiple spectrum peaks, in- [39]. Furthermore, the suggested superiority of PDFF fluence of eddy currents, and the T1 difference between over relCSA for the prediction of muscle strength seems fat and water compartments [28–30, 45]. Current se- to complement previous work at the thigh region de- quences, such as the sequence used in this study, take rived from patients with pathology, demonstrating nega- these influencing factors into account. Regarding the tive correlations between the PDFF and muscle strength muscle strength measurements, there is a need of famil- and indicating that muscle fat composition rather than iarisation sessions prior to the testing visit to ensure to muscle size correlates with knee extensor strength [5, 7– collect valid data. The rotational dynamometer testing 9, 40, 41]. procedure as well as the human ability to develop the It is important to now have evidence of correlations real maximum of isometric strength are complex so that and predictions of PDFF and muscle strength also in more visits are necessary to gain stable values of MVIC. healthy subjects as they clearly differ from patients with Therefore, all participants performed at least one initial musculoskeletal disorders, metabolic diseases, or NMD training session. regarding the ranges of PDFF measurements. PDFF There are limitations to this study that we acknowledge. values are generally much lower, and the produced First, the comparatively small sample size. Future studies strength values are higher in healthy subjects [1–9]. may enrol a larger sample size, which can be particularly Knowledge about such correlations and our prediction realised by multicentre approaches using pre-existing and model in muscles that are not or not yet pathologically state-of-the-art imaging collected in joint databases [2]. Sec- fatty infiltrated allows the PDFF to become a biomarker ond, future studies may add magnetic resonance spectros- and to potentially facilitate early treatment protocols or copy to explore the distribution of lipids within thigh to arrange counteracting individual interventions such as muscles and distinctly quantify the intra- and extra- changes in lifestyle or specific training programmes with myocellular lipid levels [13]. Third, our approach of ROI individual adapted physical activities in order to main- placement was restricted to segmenting the ten most cen- tain or improve muscle function. Former studies dealt tral slices whilst previous investigations used larger exten- with the clinical evaluation of the knee joint agonist- sion or semi-automated algorithms [2, 22, 46–48]. antagonist relationship [35, 37, 42–44]. Concerning knee However, sinceweonlyincludedhealthysubjectswho were dysfunctions and knee joint injuries, like ruptures of the characterised by rather homogeneous fat distributions, a anterior cruciate ligament, there is a high importance of manual segmentation approach of only representative slices the quadriceps-ischiocrural ratio. In our study, the con- seems to be justified [22, 46]. Fourth, concerning the meth- nection of PDFF, relCSA, and relMVIC was as follows: odical way of strength measurement, there is a need of in- in both genders, ischiocrural muscles were more infil- cluding physiological CSA, angle of pennation, and muscle trated by fat than quadriceps muscles, had only about volume in future studies as conclusive, modifying strength one third of the quadriceps relCSA, and produced about predictors and to integrate these values in prediction half of the quadriceps relMVIC. These insights pave the models [36, 49]. Fifth, the present study did not enrol pa- way for further research to prove the role of PDFF tients suffering from musculoskeletal disorders, metabolic Inhuber et al. European Radiology Experimental (2019) 3:45 Page 9 of 10 diseases, or NMD whilst insights in varying muscle quality reference database MyoSegmenTUM. PLoS One 13:e0198200. https://doi. org/10.1371/journal.pone.0198200 of different top-performing athletes are still missing. 3. Grimm A, Meyer H, Nickel MD et al (2018) Repeatability of Dixon magnetic In conclusion, we observed correlations between the resonance imaging and magnetic resonance spectroscopy for quantitative PDFF and relCSA and relMVIC at the thigh by quantita- muscle fat assessments in the thigh. J Cachexia Sarcopenia Muscle 9:1093– 1100. https://doi.org/10.1002/jcsm.12343 tive MRI and precise measurements by a rotational 4. Grimm A, Meyer H, Nickel MD et al (2018) Evaluation of 2-point, 3-point, dynamometer. In contrast to relCSA, the PDFF of the and 6-point Dixon magnetic resonance imaging with flexible echo timing quadriceps and ischiocrural muscles was significantly as- for muscle fat quantification. Eur J Radiol 103:57–64. https://doi.org/10.1016/ j.ejrad.2018.04.011 sociated bilaterally with the relMVIC in extension and 5. Kumar D, Karampinos DC, MacLeod TD et al (2014) Quadriceps intramuscular flexion. Thus, chemical shift encoding-based water-fat fat fraction rather than muscle size is associated with knee osteoarthritis. MRI can provide important information and may poten- Osteoarthritis Cartilage 22:226–234. https://doi.org/10.1016/j.joca.2013.12.005 6. Karampinos DC, Baum T, Nardo L et al (2012) Characterization of the tially track early changes in muscles not pathologically regional distribution of skeletal muscle adipose tissue in type 2 diabetes fatty infiltrated. This might help to initiate early, indivi- using chemical shift-based water/fat separation. J Magn Reson Imaging 35: dualised training protocols with specific physical activ- 899–907. https://doi.org/10.1002/jmri.23512 7. Horvath JJ, Austin SL, Case LE et al (2015) Correlation between quantitative ities in order to maintain or improve muscle function. whole-body muscle magnetic resonance imaging and clinical muscle weakness in Pompe disease. Muscle Nerve 51:722–730. https://doi.org/10. Abbreviations 1002/mus.24437 BMI: Body mass index; CSA: Cross-sectional area; MRI: Magnetic resonance 8. Willis TA, Hollingsworth KG, Coombs A et al (2013) Quantitative muscle MRI imaging; MVIC: Maximum voluntary isometric contraction; as an assessment tool for monitoring disease progression in LGMD2I: a NMD: Neuromuscular diseases; PDFF: Proton density fat fraction; multicentre longitudinal study. PLoS One 8:e70993. https://doi.org/10.1371/ relCSA: Relative CSA; RelMVIC: Relative MVIC; ROI: Region of interest; TE: Echo journal.pone.0070993 time; TR: Repetition time 9. Willis TA, Hollingsworth KG, Coombs A et al (2014) Quantitative magnetic resonance imaging in limb-girdle muscular dystrophy 2I: a multinational Authors’ contributions cross-sectional study. PLoS One 9:e90377. https://doi.org/10.1371/journal. SI, TB, and FK conceived the experiment. SI, SS, MD, and CK conducted the pone.0090377 experiment. SI, NS, EB, DCK, JSK, TB, and AS analysed the acquired data, 10. Kubota J, Ono T, Araki M, Torii S, Okuwaki T, Fukubayashi T (2007) Non- performed the statistical analysis, and interpreted the results. All authors read uniform changes in magnetic resonance measurements of the and approved the final manuscript. semitendinosus muscle following intensive eccentric exercise. Eur J Appl Physiol 101:713–720. https://doi.org/10.1007/s00421-007-0549-x Funding 11. Prior BM, Jayaraman RC, Reid RW et al (2001) Biarticular and monoarticular The project was funded by the European Research Council (grant agreement muscle activation and injury in human quadriceps muscle. Eur J Appl No 677661—ProFatMRI and grant agreement No 637164—iBack), TUM Physiol 85:185–190. https://doi.org/10.1007/s004210100434 Faculty of Medicine KKF grant H01, and Philips Healthcare. 12. Larsen RG, Ringgaard S, Overgaard K (2007) Localization and quantification of muscle damage by magnetic resonance imaging following step exercise Availability of data and materials in young women. Scand J Med Sci Sports 17:76–83. https://doi.org/10.1111/ The datasets used and analysed during the current study are available from j.1600-0838.2006.00525.x the corresponding author on request. 13. Baum T, Cordes C, Dieckmeyer M et al (2016) MR-based assessment of body fat distribution and characteristics. Eur J Radiol 85:1512–1518. https://doi. Ethics approval and consent to participate org/10.1016/j.ejrad.2016.02.013 The study was approved by the institutional review board and conducted in 14. Wernbom M, Augustsson J, Thomee R (2007) The influence of frequency, accordance with the Declaration of Helsinki. intensity, volume and mode of strength training on whole muscle cross- sectional area in humans. Sports Med 37:225–264. https://doi.org/10.2165/ Consent for publication 00007256-200737030-00004 All subjects gave written informed consent on MRI examinations, isometric 15. Jones EJ, Bishop PA, Woods AK, Green JM (2008) Cross-sectional area and strength measurements, and publication of identifying images prior to muscular strength: a brief review. Sports Med 38:987–994. https://doi.org/10. participation in the study. 2165/00007256-200838120-00003 16. Yamauchi K, Yoshiko A, Suzuki S et al (2017) Muscle atrophy and recovery of Competing interests individual thigh muscles as measured by magnetic resonance imaging scan The authors declare that they have no competing interests. during treatment with cast for ankle or foot fracture. J Orthop Surg (Hong Kong) 25:2309499017739765. https://doi.org/10.1177/2309499017739765 Author details 17. Hu HH, Kan HE (2013) Quantitative proton MR techniques for measuring fat. Department of Sport and Health Sciences, Technische Universität München, NMR Biomed 26:1609–1629. https://doi.org/10.1002/nbm.3025 Georg-Brauchle-Ring 60/62, 80992 Munich, Germany. Department of 18. Karampinos DC, Yu H, Shimakawa A, Link TM, Majumdar S (2011) T (1)- Diagnostic and Interventional Neuroradiology, Klinikum rechts der Isar, corrected fat quantification using chemical shift-based water/fat separation: Technische Universität München, Ismaninger Str. 22, 81675 Munich, Germany. application to skeletal muscle. Magn Reson Med 66:1312–1326. https://doi. Department of Diagnostic and Interventional Radiology, Klinikum rechts der org/10.1002/mrm.22925 Isar, Technische Universität München, Ismaninger Str. 22, 81675 Munich, 19. Reeder SB, Hu HH, Sirlin CB (2012) Proton density fat-fraction: a Germany. standardized MR-based biomarker of tissue fat concentration. J Magn Reson Imaging 36:1011–1014. https://doi.org/10.1002/jmri.23741 Received: 7 May 2019 Accepted: 13 September 2019 20. Burakiewicz J, Sinclair CDJ, Fischer D, Walter GA, Kan HE, Hollingsworth KG (2017) Quantifying fat replacement of muscle by quantitative MRI in muscular dystrophy. J Neurol 264:2053–2067. https://doi.org/10.1007/ References s00415-017-8547-3 1. Dahlqvist JR, Vissing CR, Thomsen C, Vissing J (2014) Severe paraspinal muscle involvement in facioscapulohumeral muscular dystrophy. Neurology 21. McGregor RA, Cameron-Smith D, Poppitt SD (2014) It is not just muscle 83:1178–1183. https://doi.org/10.1212/wnl.0000000000000828 mass: a review of muscle quality, composition and metabolism during 2. Schlaeger S, Freitag F, Klupp E et al (2018) Thigh muscle segmentation of ageing as determinants of muscle function and mobility in later life. Longev chemical shift encoding-based water-fat magnetic resonance images: the Healthspan 3:9. https://doi.org/10.1186/2046-2395-3-9 Inhuber et al. European Radiology Experimental (2019) 3:45 Page 10 of 10 22. Baum T, Inhuber S, Dieckmeyer M et al (2016) Association of quadriceps muscle 42. Aagaard P, Simonsen EB, Beyer N, Larsson B, Magnusson P, Kjaer M (1997) fat with isometric strength measurements in healthy males using chemical shift Isokinetic muscle strength and capacity for muscular knee joint stabilization encoding-based water-fat magnetic resonance imaging. J Comput Assist Tomogr in elite sailors. Int J Sports Med 18:521–525. https://doi.org/10.1055/s-2007- 40:447–451. https://doi.org/10.1097/rct.0000000000000374 972675 23. Edmunds K, Gíslason M, Sigurðsson S et al (2018) Advanced quantitative 43. Holcomb WR, Rubley MD, Lee HJ, Guadagnoli MA (2007) Effect of hamstring- methods in correlating sarcopenic muscle degeneration with lower emphasized resistance training on hamstring: quadriceps strength ratios. J extremity function biometrics and comorbidities. PLoS One 13:e0193241. Strength Cond Res 21:41–47. https://doi.org/10.1519/R-18795.1 https://doi.org/10.1371/journal.pone.0193241 44. Mau-Moeller A, Gube M, Felser S et al (2019) Intrarater reliability of muscle strength and hamstring to quadriceps strength imbalance ratios during 24. Engelke K, Museyko O, Wang L, Laredo JD (2018) Quantitative analysis of concentric, isometric, and eccentric maximal voluntary contractions using skeletal muscle by computed tomography imaging-state of the art. J the isoforce dynamometer. Clin J Sport Med 29:69–77. https://doi.org/10. Orthop Translat 15:91–103. https://doi.org/10.1371/journal.pone.0193241 1097/jsm.0000000000000493 25. Harris-Love MO, Avila NA, Adams B et al (2018) The comparative 45. Yu H, Shimakawa A, Hines CD et al (2011) Combination of complex-based and associations of ultrasound and computed tomography estimates of muscle magnitude-based multiecho water-fat separation for accurate quantification of quality with physical performance and metabolic parameters in older men. fat-fraction. Magn Reson Med 66:199–206. https://doi.org/10.1002/mrm.22840 J Clin Med 7:340. https://doi.org/10.3390/jcm7100340 46. Barnouin Y, Butler-Browne G, Voit T et al (2014) Manual segmentation of 26. Guedes DP, Lopes CC, Guedes JERP (2005) Reprodutibilidade e validade do individual muscles of the quadriceps femoris using MRI: a reappraisal. J Questionário Internacional de Atividade Física em adolescentes. Rev Bras Magn Reson Imaging 40:239–247. https://doi.org/10.1002/jmri.24370 Med Esporte 11:151–158. https://doi.org/10.1590/S1517-86922005000200011 47. Karlsson A, Rosander J, Romu T et al (2015) Automatic and quantitative 27. Kurtze N, Rangul V, Hustvedt BE (2008) Reliability and validity of the assessment of regional muscle volume by multi-atlas segmentation using international physical activity questionnaire in the Nord-Trondelag health whole-body water-fat MRI. J Magn Reson Imaging 41:1558–1569. https://doi. study (HUNT) population of men. BMC Med Res Methodol 8:63. https://doi. org/10.1002/jmri.24726 org/10.1186/1471-2288-8-63 48. Morrow JM, Sinclair CD, Fischmann A et al (2014) Reproducibility, and age, 28. Liu CY, McKenzie CA, Yu H, Brittain JH, Reeder SB (2007) Fat quantification body-weight and gender dependency of candidate skeletal muscle MRI with IDEAL gradient echo imaging: correction of bias from T (1) and noise. outcome measures in healthy volunteers. Eur Radiol 24:1610–1620. https:// Magn Reson Med 58:354–364. https://doi.org/10.1002/mrm.21301 doi.org/10.1007/s00330-014-3145-6 29. Bydder M, Yokoo T, Hamilton G et al (2008) Relaxation effects in the 49. Maughan RW, Weir J (1984) Muscle strength and cross-sectional area in quantification of fat using gradient echo imaging. Magn Reson Imaging 26: man: a comparison of strength-trained and untrained subjects. Br J Sports 347–359. https://doi.org/10.1016/j.mri.2007.08.012 Med 18:149–157. https://doi.org/10.1136/bjsm.18.3.149 30. Yu H, Shimakawa A, McKenzie CA, Brodsky E, Brittain JH, Reeder SB (2008) Multiecho water-fat separation and simultaneous R2* estimation with multifrequency fat spectrum modeling. Magn Reson Med 60:1122–1134. Publisher’sNote https://doi.org/10.1002/mrm.21737 Springer Nature remains neutral with regard to jurisdictional claims in 31. Alangari AS, Al-Hazzaa HM (2004) Normal isometric and isokinetic peak published maps and institutional affiliations. torques of hamstring and quadriceps muscles in young adult Saudi males. Neurosciences (Riyadh) 9:165–170. 23377421 32. Dirnberger J, Wiesinger H-P, Kösters A, Müller E (2012) Reproducibility for isometric and isokinetic maximum knee flexion and extension measurements using the IsoMed 2000-dynamometer. Isokinet Exerc Sci 20: 149–153. https://doi.org/10.3233/ies-2012-0451 33. Maffiuletti NA, Jubeau M, Agosti F, De Col A, Sartorio A (2008) Quadriceps muscle function characteristics in severely obese and nonobese adolescents. Eur J Appl Physiol 103:481–484. https://doi.org/10.1007/s00421-008-0737-3 34. Onishi H, Yagi R, Oyama M, Akasaka K, Ihashi K, Handa Y (2002) EMG-angle relationship of the hamstring muscles during maximum knee flexion. J Electromyogr Kinesiol 12:399–406. 12223173 35. Bollinger R, Bubeck D, Sialis I, Kramer D, Lehmann E, Alt W (2017) Relationship between muscle volume and strength asymmetry in elite athletes: investigation of the quadriceps femoris and hamstring muscles. Paper presented at the 35th conference of the International Society of Biomechanics in sports, Cologne, Germany, 14-18 June 2017 36. Blazevich AJ, Coleman DR, Horne S, Cannavan D (2009) Anatomical predictors of maximum isometric and concentric knee extensor moment. Eur J Appl Physiol 105:869–878. https://doi.org/10.1007/s00421-008-0972-7 37. Ikai M, Fukunaga T (1968) Calculation of muscle strength per unit cross- sectional area of human muscle by means of ultrasonic measurement. Int Z Angew Physiol 26:26–32 https://doi.org/10.1007/bf00696087 38. Maughan RW, Watson JS, Weir J (1983) Strength and cross-sectional area of human skeletal muscle. J Physiol 338:37–49. https://doi.org/10.1113/jphysiol. 1983.sp014658 39. Schlaeger S, Inhuber S, Rohrmeier A et al (2019) Association of paraspinal muscle water-fat MRI-based measurements with isometric strength measurements. Eur Radiol 29:599–608. https://doi.org/10.1007/s00330-018- 5631-8 40. Heymsfield SB, Gonzalez MC, Lu J, Jia G, Zheng J (2015) Skeletal muscle mass and quality: evolution of modern measurement concepts in the context of sarcopenia. Proc Nutr Soc 74:355–366. https://doi.org/10.1017/ S0029665115000129 41. Lev MH, Farkas J, Rodriguez VR et al (2001) CT angiography in the rapid triage of patients with hyperacute stroke to intraarterial thrombolysis: accuracy in the detection of large vessel thrombus. J Comput Assist Tomogr 25:520–528. https://doi.org/10.1097/00004728-200107000-00003 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Radiology Experimental Springer Journals

Associations of thigh muscle fat infiltration with isometric strength measurements based on chemical shift encoding-based water-fat magnetic resonance imaging

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/associations-of-thigh-muscle-fat-infiltration-with-isometric-strength-ssHxtEqtmq

References (51)

Publisher
Springer Journals
Copyright
Copyright © 2019 by The Author(s)
Subject
Medicine & Public Health; Imaging / Radiology; Diagnostic Radiology; Interventional Radiology; Neuroradiology; Ultrasound; Internal Medicine
eISSN
2509-9280
DOI
10.1186/s41747-019-0123-4
Publisher site
See Article on Publisher Site

Abstract

Background: Assessment of the thigh muscle fat composition using magnetic resonance imaging (MRI) can provide surrogate markers in subjects suffering from various musculoskeletal disorders including knee osteoarthritis or neuromuscular diseases. However, little is known about the relationship with muscle strength. Therefore, we investigated the associations of thigh muscle fat with isometric strength measurements. Methods: Twenty healthy subjects (10 females; median age 27 years, range 22–41 years) underwent chemical shift encoding-based water-fat MRI, followed by bilateral extraction of the proton density fat fraction (PDFF) and calculation of relative cross-sectional area (relCSA) of quadriceps and ischiocrural muscles. Relative maximum voluntary isometric contraction (relMVIC) in knee extension and flexion was measured with a rotational dynamometer. Correlations between PDFF, relCSA, and relMVIC were evaluated, and multivariate regression was applied to identify significant predictors of muscle strength. Results: Significant correlations between the PDFF and relMVIC were observed for quadriceps and ischiocrural muscles bilaterally (p = 0.001 to 0.049). PDFF, but not relCSA, was a statistically significant (p = 0.001 to 0.049) predictor of relMVIC in multivariate regression models, except for left-sided relMVIC in extension. In this case, PDFF (p = 0.005) and relCSA (p = 0.015) of quadriceps muscles significantly contributed to the statistical model with R = 0.548. adj Conclusion: Chemical shift encoding-based water-fat MRI could detect changes in muscle composition by quantifying muscular fat that correlates well with both extensor and flexor relMVIC of the thigh. Our results help to initiate early, individualised treatments to maintain or improve muscle function in subjects who do not or not yet show pathological fatty muscle infiltration. Keywords: Healthy volunteers, Magnetic resonance imaging, Muscle contraction (isometric), Muscle strength, Thigh * Correspondence: stephanie.inhuber@tum.de Department of Sport and Health Sciences, Technische Universität München, Georg-Brauchle-Ring 60/62, 80992 Munich, Germany Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Inhuber et al. European Radiology Experimental (2019) 3:45 Page 2 of 10 Key points CSA is not clear. To confirm a close relationship between Magnetic resonance imaging detects changes in thigh muscle PDFF and strength is clinically important. muscle composition by quantifying muscular fat. Thethigh musclesareoneofthe largestmusclegroupsof Muscular fat correlates well with extensor and flexor the human body and thus important target muscles to diag- strength at the thigh. nose and monitor different diseases affecting local and Muscular fat, not cross-sectional area, can predict whole-body muscle pathologies such as knee osteoarthritis, muscle strength in thigh muscles. NMD, sarcopenia, and tumour cachexia. Detecting and un- The interaction between muscular fat and strength derstanding changes in thigh muscle quality, which are could become the basis for a biomarker for muscle based on muscle fat compositions that correlate with quality and function. muscle strength, could support initiating early and indivi- dualised treatment protocols to maintain or improve Background muscle function prior to clinical diagnosis. Therefore, we Previous research has demonstrated alterations in fat investigated the association of thigh muscle fat with isomet- composition of thigh muscles due to various pathological ric strength measurement using chemical shift encoding- conditions primarily when they are already in a chronic based water-fat MRI and a rotational dynamometer in stage, including musculoskeletal disorders, metabolic dis- healthy volunteers. The hypothesis is that thigh muscle eases, and neuromuscular diseases (NMD) [1–9]. Further- PDFF improves the prediction of muscle strength measure- more, physical exercise and training have shown to entail ments beyond muscle CSA. measurable changes in the thigh musculature [10–12]. Such alterations can nowadays be captured non-invasively Methods by means of imaging, with magnetic resonance imaging Subjects (MRI) being at the forefront particularly thanks to the Twenty healthy volunteers (10 females; median age 27 ability to perform qualitative and quantitative assessments years, range 22–41 years) were recruited. Age between 20 of the human body fat composition in vivo [13]. and 45 years and body mass index (BMI) between 23 and For the purpose of assessing muscle fat composition, 33 kg/m were defined as inclusion criteria to reflect the anatomical T1- and T2-weighted MRI is conventionally average population with a rather broad BMI range (median applied [13]. Based on anatomical imaging, the cross- BMI of the study population 26.7 kg/m ,range 22.2–31.8 sectional area (CSA) of muscles can be calculated, which kg/m ). Completing the international physical activity ques- can be used as a structural measure of muscle hyper- tionnaire in short-form ensured that all subjects had a trophy or atrophy [14–16]. More advanced MRI-based moderate level of physical activity (referring to the scoring methods including magnetic resonance spectroscopy protocol: 600–1500 metabolic equivalent of task-min/week) and chemical shift encoding-based water-fat imaging en- [26, 27]. Exclusion criteria were (1) any history of high- able the extraction of parameters like the proton density performance sports, (2) any history of metabolic diseases, fat fraction (PDFF) [13, 17]. In contrast to conventional NMD, previous knee or thigh muscle injuries, (3) general T1- and T2-weighted MRI, chemical shift encoding- contraindications for MRI (e.g.,cochlearimplants),and (4) based water-fat MRI allows for robust quantitative and, implanted foreign bodies at the level of the upper leg. thus, more objective evaluation of muscle composition MRI and strength measurements were scheduled whilst simultaneously enabling anatomical assessment within 5 days to avoid any mismatch between the mea- [18–20]. In this context, thigh muscles represent a re- sured strength and fat parameters as obtained by MRI- gion of high interest for MRI investigation thanks to related analyses. In case that both parts of the measure- good magnetic field homogeneity, minimum motion ar- ment took place on the same day, MRI was carried out tefacts, and the knowledge about disease-characteristic before strength measurements with the intent to not features of some pathologies in these muscles. Addition- capture potential modifications in muscular structures ally, it is possible to perform precise strength measure- due to strength measurement. ments at thigh muscles, thus allowing for direct links to The study was approved by the local institutional review changes in quantitative MRI to provide insights into board and conducted in accordance with the Declaration muscle quality and (dys-)function [21]. of Helsinki. All subjects gave written informed consent on However, evidence of association between thigh MRI examinations, isometric strength measurements, and muscle fat composition and muscle strength is scarce publication of identifying images prior to participation in [22–25]. Inverse relationships were reported for the the study. Data are accessible upon request. PDFF and strength at the thigh in patients with NMD [1, 8]. However, whether the PDFF of extensor and MRI flexor thigh muscles improves the prediction of thigh Subjects underwent 3-T MRI of the bilateral thigh mus- isometric strength in extension and flexion beyond the cles (Ingenia, Philips Healthcare, Best, The Netherlands). Inhuber et al. European Radiology Experimental (2019) 3:45 Page 3 of 10 Scanning was performed in supine position using a multi-echo mDIXON fat quantification routine of the built-in-the-table 12-channel posterior coil and a 16- vendor. The PDFF maps were computed as the ratio of channel anterior coil. the fat signal over the sum of fat and water signals. The An axially-prescribed, six-echo three-dimensional spoiled axial PDFF maps of both stacks of each subject were gradient-echo sequence was acquired for chemical shift stored for segmentation. encoding-based water-fat separation with the following pa- rameters: repetition time (TR)/echo time (TE) min/ΔTE = Segmentation 6.4/1.1/0.8 ms; field of view 220 × 401 × 252 mm ;acquisi- Segmentations of the quadriceps femoris muscles and tion matrix 68 × 150, voxel size 3.2 × 2.0 × 4.0 mm ;fre- ischiocrural muscles of both sides were performed quency encoding direction left-to-right L/R, no sensitivity within the ten most central slices depicting the thigh encoding. Scan time was 1 min and 25 s per stack. The six muscles (Fig. 1). Segmentations were done on the PDFF echoes were acquired in a single TR using non-flyback (bi- maps using MITK (http://mitk.org/wiki/The_Medical_ polar) read-out gradients with two axial stacks being ob- Imaging_Interaction_Toolkit_(MITK); German Cancer tained consecutively to cover the entire thigh region from Research Center, Division of Medical and Biological In- the hip down to the superior edge of the patella. A flip formatics, Medical Imaging Interaction Toolkit, Heidel- angle of 3° was used to minimise T1 bias effects [18, 28]. berg, Germany). The thigh muscles were manually segmented on the first, fifth, and tenth slice. Polygonal Imaging-based fat quantification regions of interest (ROIs) were carefully placed in these The imaging data were first processed by applying a axial slices. We then used the two-dimensional phase error correction and a complex-based water-fat interpolation tool of MITK to obtain segmentations of decomposition considering a pre-calibrated seven-peak the remaining axial slices. These segmentations were fat spectrum and a single T2* [29, 30]. We used the manually corrected. The ROIs were placed at the outer Fig. 1 Chemical shift encoding-based water-fat magnetic resonance imaging (MRI) and placement of regions of interest (ROIs). a Representative proton density fat fraction (PDFF) map. b PDFF map with superimposition of manually segmented muscle compartments defined as ROIs: (1) right quadriceps muscle, (2) left quadriceps muscle, (3) right ischiocrural muscles, and (4) left ischiocrural muscles. The red lines around the thigh represent the segmentation of the entire thigh contour Inhuber et al. European Radiology Experimental (2019) 3:45 Page 4 of 10 muscle contour whilst carefully avoiding the inclusion of subcutaneous fat or muscle fat interfaces [2, 22]. Subse- quent to ROI placements, the CSA (in mm ) and PDFF (in %) of the quadriceps and ischiocrural muscles were ex- tracted separately for both sides by averaging the respect- ive values obtained from the ten consecutive slices per side. Additionally, the entire thigh contour was segmented in the same ten slices (Fig. 1), followed by calculation of the mean thigh CSA per subject. A relative CSA (relCSA, in %) of quadriceps and ischiocrural muscles was deter- mined by dividing the respective muscle CSA by the entire thigh CSA. All segmentations were performed by a radi- ologist with 8 years of experience in musculoskeletal im- aging. Good reproducibility of measurements following this approach was previously reported [22]. Isometric muscle strength measurements The maximum voluntary isometric contraction (MVIC) in single-joint knee extension and flexion was measured at the thigh bilaterally with a rotational dynamometer (IsoMed 2000, D&R Ferstl GmbH, Hemau, Germany). Substantiated by measuring the isometric peak torque in Newton per Fig. 2 Setup for measurements of the maximum voluntary isometric metre (in N*m ), the MVIC was produced in knee exten- contraction (MVIC) with a rotational dynamometer sion at 60° and knee flexion at 35°, which are the joint an- gles with the ideal muscular strength-length relationship to anticipate the real MVIC [31–34]. At the beginning of the transfer the measured values of maximum isometric measuring visit, the isokinetic rotational dynamometer was torque to the software (proEMG, Prophysics AG, calibrated exactly on subjects’ individual body dimension. Switzerland). The value the tested leg produces whilst Subjects were seated in an upright sitting position and were sitting in a resting position caused by the gravity was fixed by hip and shoulder belts as well as shoulder pads, measured in each testing step. The measured absolute and two adjustable straps were used to fix the leg at the extension and flexion MVICs (in N*m ) were adjusted for pad of the lever arm in the correct measuring position the individual BMI to obtain a relative MVIC (relMVIC, (Fig. 2). in N*m /kg) for the left and right thigh muscles. Following the individual calibrating, the subjects had to perform a standardised warm-up exercise. After 10 Reproducibility of muscle strength measurements min on a cycling ergometer at 70–80 rpm (revolutions All participants performed at least one initial training per minute) to activate the cardiovascular system, the session. The visit’s aim was to become familiar with the volunteers were seated in the rotational dynamometer measuring procedure and to train to afford and activate where ten dynamic repetitions with moderate intensity the maximum isometric strength. To ensure the repro- were performed. In the fixed measuring position, the ducibility of strength measurements, each subject per- subjects’ task was to extend or to flex the knee against formed at least five to eight repetitions with a MVIC the measuring pad on the back or front side of the lower and 3 min of rest in between. A verifiable increase and leg with the individual maximum contraction of quadri- decrease of the measured values was assured as the de- ceps or ischiocrural muscles. MVIC of each direction of cisive factor for a potentially needed second training visit movement (extension/flexion) was collected three times or the testing appointment. There were at least 3 days of with 3 min of recovery in between, and the best value of rest between the visits and subjects were instructed to muscle flexion and extension maximum isometric torque come to the measurements totally recovered (no physical (in N*m ) was respectively taken for data analysis [2, activity the 2 days before visits). 22]. The choice of the starting leg and the direction of movement were randomised. Statistical analyses SPSS (version 20.0; IBM SPSS Statistics for Windows, Strength data collection method Armonk, NY, USA) was used for all statistical analyses The force transducer is located in the inside of the rota- and generation of graphs. The level of statistical signifi- tional lever arm of the used rotational dynamometer to cance was set at p < 0.05 (two-sided). The Kolmogorov- Inhuber et al. European Radiology Experimental (2019) 3:45 Page 5 of 10 Smirnov test indicated normal distribution of BMI, Muscle strength measurements PDFF, relCSA, and relMVIC values, but not of age. RelMVIC in extension was higher in males compared to First, mean ± standard deviation was calculated for PDFF, females in the left (8.0 ± 1.2 N*m /kg versus 5.9 ± 1.0 relCSA, and relMVIC, separately for male and female sub- N*m /kg, p < 0.001) and right quadriceps muscles 3 3 jects as well as for the right and left thigh muscles. Compar- (8.8 ± 1.3 N*m /kg versus 6.4 ± 0.9 N*m /kg, p < 0.001; isons of these measures between genders were performed Table 1). Similarly, males had greater relMVIC in flexion by using unpaired t-tests. Furthermore, Pearson correlation than females in the left (4.0 ± 0.6 N*m /kg versus 3.0 ± coefficients (r) were calculated between PDFF, relCSA, and 0.5 N*m /kg, p < 0.001) and right ischiocrural muscles 3 3 relMVIC. Using these parameters, multivariate regression (4.3 ± 0.6 N*m /kg versus 2.7 ± 0.7 N*m /kg, p < 0.001; models were calculated to identify significant predictors of Table 1). Quadriceps muscles showed approximately thigh muscle strength. Parameters were included in the re- twice as much relMVIC than ischiocrural muscles at gression models for p < 0.05. both thighs in males (mean quadriceps 8.4 ± 1.2 N*m /kg versus mean ischiocrural 4.1 ± 0.6 N*m /kg) and females (mean quadriceps 6.2 ± 1.0 N*m /kg versus mean ischio- Results crural 2.9 ± 0.6 N*m /kg; Table 1). Proton density fat fraction and cross-sectional area measurements Mean muscle PDFF was lower in males than females in the Correlations and multivariate regression models left and right quadriceps muscles (2.7 ± 1.3% versus 3.6 ± Significant correlations were revealed between the relMVIC 1.3%, p = 0.162; 1.7 ± 1.3% versus 2.7 ± 1.3%, p = 0.105) and in extension and the PDFF of the left (r = -0.649, p =0.002) ischiocrural muscles (3.2 ± 1.6% versus 4.6 ± 2.0%, p =0.091; and right quadriceps muscle (r = -0.612, p = 0.004; Table 2, 2.6 ± 1.9% versus 4.9 ± 2.4%, p = 0.025; Table 1). There were Fig. 3). RelMVIC in flexion was correlated significantly with no significant differences in age or BMI between males and the PDFF of the left (r = -0.446, p = 0.049) and right ischio- females (p = 0.280 and 0.684, respectively). crural muscles (r = -0.676, p =0.001; Table 2,Fig. 3). Males showed greater relCSA than females in the quad- Furthermore, the relMVIC in extension was signifi- riceps muscles (26.5 ± 4.0% versus 21.3 ± 4.1%, p =0.010; cantly associated with the relCSA of the left (r = 0.585, 25.8 ± 4.2% versus 21.8 ± 3.7%, p =0.035) and ischiocrural p = 0.007) and right quadriceps muscle (r = 0.448, p = muscles at both sides (8.3 ± 2.4% versus 6.0 ± 2.4%, p = 0.048; Table 2, Fig. 4). There was no significant correl- 0.045; 7.5 ± 3.0% versus 6.6 ± 1.6%, p =0.418; Table 1). ation found between the relMVIC in flexion and relCSA Quadriceps muscles had a bigger relCSA than ischiocrural of the ischiocrural muscles for the left and right sides muscles for males and females (Table 1). (Table 2, Fig. 4). PDFF, but not relCSA, was a statistically significant (p = 0.001 to 0.049) predictor of relMVIC in multivariate Table 1 Proton density fat fraction (PDFF), relative cross- regression models, except for left-sided relMVIC in ex- sectional area (relCSA), and relative maximum voluntary tension. In this case, PDFF (p = 0.005) and relCSA (p = isometric contraction (relMVIC) in extension and flexion 0.015) of the quadriceps muscles contributed signifi- Males Females p value cantly to the statistical model with R = 0.548. adj PDFF and relCSA PDFF quadriceps (%) Left 2.7 ± 1.3 3.6 ± 1.3 0.162 Right 1.7 ± 1.3 2.7 ± 1.3 0.105 Table 2 Correlation between the proton density fat fraction PDFF ischiocrural (%) Left 3.2 ± 1.6 4.6 ± 2.0 0.091 (PDFF) or relative cross-sectional area (relCSA) and relative Right 2.6 ± 1.9 4.9 ± 2.4 0.025 maximum voluntary isometric contraction (relMVIC) in extension relCSA quadriceps (%) Left 26.5 ± 4.0 21.3 ± 4.1 0.010 and flexion (males and females together) Right 25.8 ± 4.2 21.8 ± 3.7 0.035 relMVIC relCSA relCSA ischiocrural (%) Left 8.3 ± 2.4 6.0 ± 2.4 0.045 Left Right Left Right Right 7.5 ± 3.0 6.6 ± 1.6 0.418 Extension PDFF r -0.649 -0.612 −0.284 − 0.308 p value 0.002 0.004 0.225 0.186 RelMVIC in extension and flexion relCSA r 0.585 0.448 –– relMVIC in extension (N*m /kg) Left 8.0 ± 1.2 5.9 ± 1.0 < 0.001 p value 0.007 0.048 Right 8.8 ± 1.3 6.4 ± 0.9 < 0.001 Flexion PDFF r -0.446 -0.676 -0.125 -0.173 p value 0.049 0.001 0.599 0.465 relMVIC in flexion (N*m /kg) Left 4.0 ± 0.6 3.0 ± 0.5 < 0.001 relCSA r 0.238 0.207 –– Right 4.3 ± 0.6 2.7 ± 0.7 < 0.001 p value 0.312 0.380 Data are expressed as mean ± standard deviation Inhuber et al. European Radiology Experimental (2019) 3:45 Page 6 of 10 Fig. 3 Correlation between the proton density fat fraction (PDFF) and relative maximum voluntary isometric contraction (relMVIC). Plots showing the association between the left or right relMVIC in extension or flexion (in N*m /kg) and the PDFF (in %) of the left or right quadriceps or ischiocrural muscles Discussion muscles as part of the MyoSegmenTUM_thigh database, This study used chemical shift encoding-based water-fat with average PDFF values of 3.71% and 3.93% for the right MRI at the thigh in healthy volunteers to extract the and left quadriceps muscle as well as 4.38% and 4.44% for PDFF and relCSA of quadriceps and ischiocrural mus- the right and left ischiocrural muscles by also focusing on cles bilaterally and a rotational dynamometer to assess healthy males and females. However, a distinction be- relMVIC in extension and flexion. Muscle PDFF was a tween genders has not been achieved in their study be- better predictor of the relMVIC than the relCSA. We cause only three healthy females were included in total observed significant differences in the relCSA of ischio- [2]. Thus, the PDFF values obtained in the present study crural and quadriceps muscles, but also in the PDFF of seem to be principally in the range of previously reported ischiocrural muscles between males and females. values. Grimm et al. [3] recently reported averaged PDFF values Alterations in fat composition of thigh muscles have between 5.6 and 6.9% in healthy young males; however, in been shown in the context of diseases such as musculo- their investigation, PDFF measurements were derived skeletal disorders, metabolic diseases, and NMD [1–9]. from all thigh muscles and without distinct exclusion of Specifically, patients suffering from different types of intermuscular tissue. Specifically for the quadriceps mus- muscular dystrophy, Pompe disease, sarcopenia, osteo- cles, a previous study [22] reported a mean intramuscular arthritis, or type 2 diabetes mellitus showed increased PDFF of 4.02% in a cohort of healthy males. Schlaeger PDFF or intramuscular fat when compared to healthy et al. [2] performed the segmentation of individual thigh controls [1–9]. Although it has been hypothesised that Inhuber et al. European Radiology Experimental (2019) 3:45 Page 7 of 10 Fig. 4 Correlation between the relative cross-sectional area (relCSA) and relative maximum voluntary isometric contraction (relMVIC). Plots showing the association between the left or right relMVIC in extension or flexion (in N*m /kg) and the relative CSA (in %) of the left or right quadriceps or ischiocrural muscles increased fatty infiltration or transformation of thigh knee extension and indicated that the quadriceps inter- muscles is related to decrease in muscle strength [21, muscular adipose tissue fraction and intramuscular 23], studies correlating parameters such as the PDFF or PDFF correlate significantly with physical strength, rep- absolute or relative CSA of thigh muscles with objective resented by the MVIC. measurements of muscle strength are rare. An inverse The present study confirms this first evidence of an relationship was revealed between the PDFF and interaction between muscular fat and strength by examin- strength at the thigh in patients with NMD [1, 8]. In ing the associations between the PDFF and relMVIC at these studies, the obtained strength measurements were the thigh in healthy subjects and by showing not only sig- acquired with a handheld myometer [1, 8]; however, the nificant correlations for the PDFF of the quadriceps mus- usage of more objective devices to test muscle strength, cles but also for ischiocrural muscles. Moreover, we were such as isokinetic dynamometers that enable robust as- able to show that, in contrast to the relCSA, the PDFF of sessments of the muscle strength of specific functional the quadriceps and ischiocrural muscles were significantly muscle groups, is still largely missing to evaluate the as- associated bilaterally with relMVIC in knee extension and sociation between fat components and thigh muscle flexion. function. One previous study [22] focusing on the fat- These new insights in the interactions of fat and strength interaction of quadriceps muscles by only test- strength parameters in human muscles demonstrate the ing a small cohort of healthy subjects used an isokinetic importance of muscle quality, consisting of the individual dynamometer for isometric strength measurement in number of muscular contractile elements, the specific fatty Inhuber et al. European Radiology Experimental (2019) 3:45 Page 8 of 10 infiltration as well as the strength capacity [21]. The concerning muscle function and its role as an indicator muscle quality might be able to predict and flesh out to analyse and define muscle quality and fat-induced loss muscle (dys-)function a lot better than the relCSA does. of muscle function as well as to develop corresponding Of note, there is no significant correlation between the training programmes dealing with these ratios. As a per- relMVIC in flexion and the relCSA of the ischiocrural spective, the role of physiological CSA, muscular penna- muscles for the left and right side. The anatomical differ- tion angle, and muscle volume could complement and ence (quadriceps muscles: one-joint muscles; ischiocrural specify muscle quality [32, 35]. muscles: multi-joint muscles) as well as the variations con- Concerning methodology, chemical shift encoding- cerning the muscular structure and activation should be based water-fat MRI is confirmed as a fast method considered when interpreting this finding [35]. To specify which can be added to routine MRI protocols of the the knowledge of muscle quality and (dys-)function in the thigh region to visualise and quantify muscle quality and future, further studies are needed to include and prove the possibly forecast deficits in muscle strength and func- meaning of the muscle volume and in particular the tion, leading over to subsequent specific training pro- physiological CSA, the probably most powerful predictors grammes. In this context, PDFF derived from chemical concerning muscular strength [36–38]. shift encoding-based water-fat MRI is a more robust and Our observations seem to be in accordance with the objective approach compared to the semi-quantitative, finding of paraspinal PDFF being significantly correlated post-acquisitional analysis of conventional images [13]. with the relMVIC in extension and flexion of the trunk, However, the method requires compensation for con- whereas paraspinal mean CSA only showed significant founding factors, reflected by T2* decay, a potential correlations with relMVIC in flexion in a previous study quantification bias due to multiple spectrum peaks, in- [39]. Furthermore, the suggested superiority of PDFF fluence of eddy currents, and the T1 difference between over relCSA for the prediction of muscle strength seems fat and water compartments [28–30, 45]. Current se- to complement previous work at the thigh region de- quences, such as the sequence used in this study, take rived from patients with pathology, demonstrating nega- these influencing factors into account. Regarding the tive correlations between the PDFF and muscle strength muscle strength measurements, there is a need of famil- and indicating that muscle fat composition rather than iarisation sessions prior to the testing visit to ensure to muscle size correlates with knee extensor strength [5, 7– collect valid data. The rotational dynamometer testing 9, 40, 41]. procedure as well as the human ability to develop the It is important to now have evidence of correlations real maximum of isometric strength are complex so that and predictions of PDFF and muscle strength also in more visits are necessary to gain stable values of MVIC. healthy subjects as they clearly differ from patients with Therefore, all participants performed at least one initial musculoskeletal disorders, metabolic diseases, or NMD training session. regarding the ranges of PDFF measurements. PDFF There are limitations to this study that we acknowledge. values are generally much lower, and the produced First, the comparatively small sample size. Future studies strength values are higher in healthy subjects [1–9]. may enrol a larger sample size, which can be particularly Knowledge about such correlations and our prediction realised by multicentre approaches using pre-existing and model in muscles that are not or not yet pathologically state-of-the-art imaging collected in joint databases [2]. Sec- fatty infiltrated allows the PDFF to become a biomarker ond, future studies may add magnetic resonance spectros- and to potentially facilitate early treatment protocols or copy to explore the distribution of lipids within thigh to arrange counteracting individual interventions such as muscles and distinctly quantify the intra- and extra- changes in lifestyle or specific training programmes with myocellular lipid levels [13]. Third, our approach of ROI individual adapted physical activities in order to main- placement was restricted to segmenting the ten most cen- tain or improve muscle function. Former studies dealt tral slices whilst previous investigations used larger exten- with the clinical evaluation of the knee joint agonist- sion or semi-automated algorithms [2, 22, 46–48]. antagonist relationship [35, 37, 42–44]. Concerning knee However, sinceweonlyincludedhealthysubjectswho were dysfunctions and knee joint injuries, like ruptures of the characterised by rather homogeneous fat distributions, a anterior cruciate ligament, there is a high importance of manual segmentation approach of only representative slices the quadriceps-ischiocrural ratio. In our study, the con- seems to be justified [22, 46]. Fourth, concerning the meth- nection of PDFF, relCSA, and relMVIC was as follows: odical way of strength measurement, there is a need of in- in both genders, ischiocrural muscles were more infil- cluding physiological CSA, angle of pennation, and muscle trated by fat than quadriceps muscles, had only about volume in future studies as conclusive, modifying strength one third of the quadriceps relCSA, and produced about predictors and to integrate these values in prediction half of the quadriceps relMVIC. These insights pave the models [36, 49]. Fifth, the present study did not enrol pa- way for further research to prove the role of PDFF tients suffering from musculoskeletal disorders, metabolic Inhuber et al. European Radiology Experimental (2019) 3:45 Page 9 of 10 diseases, or NMD whilst insights in varying muscle quality reference database MyoSegmenTUM. PLoS One 13:e0198200. https://doi. org/10.1371/journal.pone.0198200 of different top-performing athletes are still missing. 3. Grimm A, Meyer H, Nickel MD et al (2018) Repeatability of Dixon magnetic In conclusion, we observed correlations between the resonance imaging and magnetic resonance spectroscopy for quantitative PDFF and relCSA and relMVIC at the thigh by quantita- muscle fat assessments in the thigh. J Cachexia Sarcopenia Muscle 9:1093– 1100. https://doi.org/10.1002/jcsm.12343 tive MRI and precise measurements by a rotational 4. Grimm A, Meyer H, Nickel MD et al (2018) Evaluation of 2-point, 3-point, dynamometer. In contrast to relCSA, the PDFF of the and 6-point Dixon magnetic resonance imaging with flexible echo timing quadriceps and ischiocrural muscles was significantly as- for muscle fat quantification. Eur J Radiol 103:57–64. https://doi.org/10.1016/ j.ejrad.2018.04.011 sociated bilaterally with the relMVIC in extension and 5. Kumar D, Karampinos DC, MacLeod TD et al (2014) Quadriceps intramuscular flexion. Thus, chemical shift encoding-based water-fat fat fraction rather than muscle size is associated with knee osteoarthritis. MRI can provide important information and may poten- Osteoarthritis Cartilage 22:226–234. https://doi.org/10.1016/j.joca.2013.12.005 6. Karampinos DC, Baum T, Nardo L et al (2012) Characterization of the tially track early changes in muscles not pathologically regional distribution of skeletal muscle adipose tissue in type 2 diabetes fatty infiltrated. This might help to initiate early, indivi- using chemical shift-based water/fat separation. J Magn Reson Imaging 35: dualised training protocols with specific physical activ- 899–907. https://doi.org/10.1002/jmri.23512 7. Horvath JJ, Austin SL, Case LE et al (2015) Correlation between quantitative ities in order to maintain or improve muscle function. whole-body muscle magnetic resonance imaging and clinical muscle weakness in Pompe disease. Muscle Nerve 51:722–730. https://doi.org/10. Abbreviations 1002/mus.24437 BMI: Body mass index; CSA: Cross-sectional area; MRI: Magnetic resonance 8. Willis TA, Hollingsworth KG, Coombs A et al (2013) Quantitative muscle MRI imaging; MVIC: Maximum voluntary isometric contraction; as an assessment tool for monitoring disease progression in LGMD2I: a NMD: Neuromuscular diseases; PDFF: Proton density fat fraction; multicentre longitudinal study. PLoS One 8:e70993. https://doi.org/10.1371/ relCSA: Relative CSA; RelMVIC: Relative MVIC; ROI: Region of interest; TE: Echo journal.pone.0070993 time; TR: Repetition time 9. Willis TA, Hollingsworth KG, Coombs A et al (2014) Quantitative magnetic resonance imaging in limb-girdle muscular dystrophy 2I: a multinational Authors’ contributions cross-sectional study. PLoS One 9:e90377. https://doi.org/10.1371/journal. SI, TB, and FK conceived the experiment. SI, SS, MD, and CK conducted the pone.0090377 experiment. SI, NS, EB, DCK, JSK, TB, and AS analysed the acquired data, 10. Kubota J, Ono T, Araki M, Torii S, Okuwaki T, Fukubayashi T (2007) Non- performed the statistical analysis, and interpreted the results. All authors read uniform changes in magnetic resonance measurements of the and approved the final manuscript. semitendinosus muscle following intensive eccentric exercise. Eur J Appl Physiol 101:713–720. https://doi.org/10.1007/s00421-007-0549-x Funding 11. Prior BM, Jayaraman RC, Reid RW et al (2001) Biarticular and monoarticular The project was funded by the European Research Council (grant agreement muscle activation and injury in human quadriceps muscle. Eur J Appl No 677661—ProFatMRI and grant agreement No 637164—iBack), TUM Physiol 85:185–190. https://doi.org/10.1007/s004210100434 Faculty of Medicine KKF grant H01, and Philips Healthcare. 12. Larsen RG, Ringgaard S, Overgaard K (2007) Localization and quantification of muscle damage by magnetic resonance imaging following step exercise Availability of data and materials in young women. Scand J Med Sci Sports 17:76–83. https://doi.org/10.1111/ The datasets used and analysed during the current study are available from j.1600-0838.2006.00525.x the corresponding author on request. 13. Baum T, Cordes C, Dieckmeyer M et al (2016) MR-based assessment of body fat distribution and characteristics. Eur J Radiol 85:1512–1518. https://doi. Ethics approval and consent to participate org/10.1016/j.ejrad.2016.02.013 The study was approved by the institutional review board and conducted in 14. Wernbom M, Augustsson J, Thomee R (2007) The influence of frequency, accordance with the Declaration of Helsinki. intensity, volume and mode of strength training on whole muscle cross- sectional area in humans. Sports Med 37:225–264. https://doi.org/10.2165/ Consent for publication 00007256-200737030-00004 All subjects gave written informed consent on MRI examinations, isometric 15. Jones EJ, Bishop PA, Woods AK, Green JM (2008) Cross-sectional area and strength measurements, and publication of identifying images prior to muscular strength: a brief review. Sports Med 38:987–994. https://doi.org/10. participation in the study. 2165/00007256-200838120-00003 16. Yamauchi K, Yoshiko A, Suzuki S et al (2017) Muscle atrophy and recovery of Competing interests individual thigh muscles as measured by magnetic resonance imaging scan The authors declare that they have no competing interests. during treatment with cast for ankle or foot fracture. J Orthop Surg (Hong Kong) 25:2309499017739765. https://doi.org/10.1177/2309499017739765 Author details 17. Hu HH, Kan HE (2013) Quantitative proton MR techniques for measuring fat. Department of Sport and Health Sciences, Technische Universität München, NMR Biomed 26:1609–1629. https://doi.org/10.1002/nbm.3025 Georg-Brauchle-Ring 60/62, 80992 Munich, Germany. Department of 18. Karampinos DC, Yu H, Shimakawa A, Link TM, Majumdar S (2011) T (1)- Diagnostic and Interventional Neuroradiology, Klinikum rechts der Isar, corrected fat quantification using chemical shift-based water/fat separation: Technische Universität München, Ismaninger Str. 22, 81675 Munich, Germany. application to skeletal muscle. Magn Reson Med 66:1312–1326. https://doi. Department of Diagnostic and Interventional Radiology, Klinikum rechts der org/10.1002/mrm.22925 Isar, Technische Universität München, Ismaninger Str. 22, 81675 Munich, 19. Reeder SB, Hu HH, Sirlin CB (2012) Proton density fat-fraction: a Germany. standardized MR-based biomarker of tissue fat concentration. J Magn Reson Imaging 36:1011–1014. https://doi.org/10.1002/jmri.23741 Received: 7 May 2019 Accepted: 13 September 2019 20. Burakiewicz J, Sinclair CDJ, Fischer D, Walter GA, Kan HE, Hollingsworth KG (2017) Quantifying fat replacement of muscle by quantitative MRI in muscular dystrophy. J Neurol 264:2053–2067. https://doi.org/10.1007/ References s00415-017-8547-3 1. Dahlqvist JR, Vissing CR, Thomsen C, Vissing J (2014) Severe paraspinal muscle involvement in facioscapulohumeral muscular dystrophy. Neurology 21. McGregor RA, Cameron-Smith D, Poppitt SD (2014) It is not just muscle 83:1178–1183. https://doi.org/10.1212/wnl.0000000000000828 mass: a review of muscle quality, composition and metabolism during 2. Schlaeger S, Freitag F, Klupp E et al (2018) Thigh muscle segmentation of ageing as determinants of muscle function and mobility in later life. Longev chemical shift encoding-based water-fat magnetic resonance images: the Healthspan 3:9. https://doi.org/10.1186/2046-2395-3-9 Inhuber et al. European Radiology Experimental (2019) 3:45 Page 10 of 10 22. Baum T, Inhuber S, Dieckmeyer M et al (2016) Association of quadriceps muscle 42. Aagaard P, Simonsen EB, Beyer N, Larsson B, Magnusson P, Kjaer M (1997) fat with isometric strength measurements in healthy males using chemical shift Isokinetic muscle strength and capacity for muscular knee joint stabilization encoding-based water-fat magnetic resonance imaging. J Comput Assist Tomogr in elite sailors. Int J Sports Med 18:521–525. https://doi.org/10.1055/s-2007- 40:447–451. https://doi.org/10.1097/rct.0000000000000374 972675 23. Edmunds K, Gíslason M, Sigurðsson S et al (2018) Advanced quantitative 43. Holcomb WR, Rubley MD, Lee HJ, Guadagnoli MA (2007) Effect of hamstring- methods in correlating sarcopenic muscle degeneration with lower emphasized resistance training on hamstring: quadriceps strength ratios. J extremity function biometrics and comorbidities. PLoS One 13:e0193241. Strength Cond Res 21:41–47. https://doi.org/10.1519/R-18795.1 https://doi.org/10.1371/journal.pone.0193241 44. Mau-Moeller A, Gube M, Felser S et al (2019) Intrarater reliability of muscle strength and hamstring to quadriceps strength imbalance ratios during 24. Engelke K, Museyko O, Wang L, Laredo JD (2018) Quantitative analysis of concentric, isometric, and eccentric maximal voluntary contractions using skeletal muscle by computed tomography imaging-state of the art. J the isoforce dynamometer. Clin J Sport Med 29:69–77. https://doi.org/10. Orthop Translat 15:91–103. https://doi.org/10.1371/journal.pone.0193241 1097/jsm.0000000000000493 25. Harris-Love MO, Avila NA, Adams B et al (2018) The comparative 45. Yu H, Shimakawa A, Hines CD et al (2011) Combination of complex-based and associations of ultrasound and computed tomography estimates of muscle magnitude-based multiecho water-fat separation for accurate quantification of quality with physical performance and metabolic parameters in older men. fat-fraction. Magn Reson Med 66:199–206. https://doi.org/10.1002/mrm.22840 J Clin Med 7:340. https://doi.org/10.3390/jcm7100340 46. Barnouin Y, Butler-Browne G, Voit T et al (2014) Manual segmentation of 26. Guedes DP, Lopes CC, Guedes JERP (2005) Reprodutibilidade e validade do individual muscles of the quadriceps femoris using MRI: a reappraisal. J Questionário Internacional de Atividade Física em adolescentes. Rev Bras Magn Reson Imaging 40:239–247. https://doi.org/10.1002/jmri.24370 Med Esporte 11:151–158. https://doi.org/10.1590/S1517-86922005000200011 47. Karlsson A, Rosander J, Romu T et al (2015) Automatic and quantitative 27. Kurtze N, Rangul V, Hustvedt BE (2008) Reliability and validity of the assessment of regional muscle volume by multi-atlas segmentation using international physical activity questionnaire in the Nord-Trondelag health whole-body water-fat MRI. J Magn Reson Imaging 41:1558–1569. https://doi. study (HUNT) population of men. BMC Med Res Methodol 8:63. https://doi. org/10.1002/jmri.24726 org/10.1186/1471-2288-8-63 48. Morrow JM, Sinclair CD, Fischmann A et al (2014) Reproducibility, and age, 28. Liu CY, McKenzie CA, Yu H, Brittain JH, Reeder SB (2007) Fat quantification body-weight and gender dependency of candidate skeletal muscle MRI with IDEAL gradient echo imaging: correction of bias from T (1) and noise. outcome measures in healthy volunteers. Eur Radiol 24:1610–1620. https:// Magn Reson Med 58:354–364. https://doi.org/10.1002/mrm.21301 doi.org/10.1007/s00330-014-3145-6 29. Bydder M, Yokoo T, Hamilton G et al (2008) Relaxation effects in the 49. Maughan RW, Weir J (1984) Muscle strength and cross-sectional area in quantification of fat using gradient echo imaging. Magn Reson Imaging 26: man: a comparison of strength-trained and untrained subjects. Br J Sports 347–359. https://doi.org/10.1016/j.mri.2007.08.012 Med 18:149–157. https://doi.org/10.1136/bjsm.18.3.149 30. Yu H, Shimakawa A, McKenzie CA, Brodsky E, Brittain JH, Reeder SB (2008) Multiecho water-fat separation and simultaneous R2* estimation with multifrequency fat spectrum modeling. Magn Reson Med 60:1122–1134. Publisher’sNote https://doi.org/10.1002/mrm.21737 Springer Nature remains neutral with regard to jurisdictional claims in 31. Alangari AS, Al-Hazzaa HM (2004) Normal isometric and isokinetic peak published maps and institutional affiliations. torques of hamstring and quadriceps muscles in young adult Saudi males. Neurosciences (Riyadh) 9:165–170. 23377421 32. Dirnberger J, Wiesinger H-P, Kösters A, Müller E (2012) Reproducibility for isometric and isokinetic maximum knee flexion and extension measurements using the IsoMed 2000-dynamometer. Isokinet Exerc Sci 20: 149–153. https://doi.org/10.3233/ies-2012-0451 33. Maffiuletti NA, Jubeau M, Agosti F, De Col A, Sartorio A (2008) Quadriceps muscle function characteristics in severely obese and nonobese adolescents. Eur J Appl Physiol 103:481–484. https://doi.org/10.1007/s00421-008-0737-3 34. Onishi H, Yagi R, Oyama M, Akasaka K, Ihashi K, Handa Y (2002) EMG-angle relationship of the hamstring muscles during maximum knee flexion. J Electromyogr Kinesiol 12:399–406. 12223173 35. Bollinger R, Bubeck D, Sialis I, Kramer D, Lehmann E, Alt W (2017) Relationship between muscle volume and strength asymmetry in elite athletes: investigation of the quadriceps femoris and hamstring muscles. Paper presented at the 35th conference of the International Society of Biomechanics in sports, Cologne, Germany, 14-18 June 2017 36. Blazevich AJ, Coleman DR, Horne S, Cannavan D (2009) Anatomical predictors of maximum isometric and concentric knee extensor moment. Eur J Appl Physiol 105:869–878. https://doi.org/10.1007/s00421-008-0972-7 37. Ikai M, Fukunaga T (1968) Calculation of muscle strength per unit cross- sectional area of human muscle by means of ultrasonic measurement. Int Z Angew Physiol 26:26–32 https://doi.org/10.1007/bf00696087 38. Maughan RW, Watson JS, Weir J (1983) Strength and cross-sectional area of human skeletal muscle. J Physiol 338:37–49. https://doi.org/10.1113/jphysiol. 1983.sp014658 39. Schlaeger S, Inhuber S, Rohrmeier A et al (2019) Association of paraspinal muscle water-fat MRI-based measurements with isometric strength measurements. Eur Radiol 29:599–608. https://doi.org/10.1007/s00330-018- 5631-8 40. Heymsfield SB, Gonzalez MC, Lu J, Jia G, Zheng J (2015) Skeletal muscle mass and quality: evolution of modern measurement concepts in the context of sarcopenia. Proc Nutr Soc 74:355–366. https://doi.org/10.1017/ S0029665115000129 41. Lev MH, Farkas J, Rodriguez VR et al (2001) CT angiography in the rapid triage of patients with hyperacute stroke to intraarterial thrombolysis: accuracy in the detection of large vessel thrombus. J Comput Assist Tomogr 25:520–528. https://doi.org/10.1097/00004728-200107000-00003

Journal

European Radiology Experimental – Springer Journals

Published: Nov 20, 2019

There are no references for this article.