Skin Strain Analysis of the Scapular Region and Wearables Design

Monitoring scapular movements is of relevance in the contexts of rehabilitation and clinical research. Among many technologies, wearable systems instrumented by strain sensors are emerging in these applications. An open challenge for the design of these systems is the optimal positioning of the sensing elements, since their response is related to the strain of the underlying substrates. This study aimed to provide a method to analyze the human skin strain of the scapular region. Experiments were conducted on five healthy volunteers to assess the skin strain during upper limb movements in the frontal, sagittal, and scapular planes at different degrees of elevation. A 6 × 5 grid of passive markers was placed posteriorly to cover the entire anatomic region of interest. Results showed that the maximum strain values, in percentage, were 28.26%, and 52.95%, 60.12% and 60.87%, 40.89%, and 48.20%, for elevation up to 90° and maximum elevation in the frontal, sagittal, and scapular planes, respectively. In all cases, the maximum extension is referred to the pair of markers placed horizontally near the axillary fold. Accordingly, this study suggests interesting insights for designing and positioning textile-based strain sensors in wearable systems for scapular movements monitoring.


Introduction
Patients suffering from shoulder musculoskeletal disorders (MSDs) may experience pain and reduced functional capacity [1,2]. The scapula, the bone linking the humerus with the clavicle, ensures proper alignment and the normal mobility of the glenohumeral and acromioclavicular joints [3]. A correct and coordinated scapular movement represents the key component in regular shoulder functionality. Alterations in scapular position and orientation, a condition known as scapular dyskinesis, characterize most shoulder MSDs, such as subacromial impingement syndrome, rotator cuff tears, frozen shoulder, or multidirectional instability [1,4]. The scapulothoracic joint is a functional sliding joint between the medial border of the scapula and the posterior thoracic ribcage, allowing the relative motion of the scapula on the thoracic surface below. The joint variables in sliding joints are the extensions between two sequential body segments [5,6].
Obtaining objective data of the scapular movements considering both different degrees of elevations and planes (e.g., frontal, sagittal, and scapular) could provide meaningful achievements in the context of rehabilitation and clinical research [7][8][9]. Recently, increasing attention has been directed toward understanding the complex scapula kinematics and providing monitoring systems that can quantify scapular movements [8,10]. To date, several axis along the scapular spine [31]. Translations include superior-inferior (elevation-depression) and mediolateral (retraction-protraction) motions of the scapulae over the posterior chest wall. Translational movements are permitted by the connection of the scapula to the axial skeleton through the clavicle [31]. Figure 1 illustrates the main scapular movements. The scapula serves as the location of various muscles' attachment [3]. Such muscles, having different sizes, functions, and depths, experience several stretching directions during upper limb movements in the different planes of the 3D space and at different degrees of elevation. Moreover, the scapula posteriorly is covered by overlying soft tissue, which in turn influences the superficial deformation of the scapular region. For all of these reasons, the skin deformations in the scapular region have dissimilarity in stretching position and magnitude during upper limbs motions. The main scapulothoracic muscles are the trapezius muscle, the serratus anterior muscle, the rhomboids, and the levator scapulae [3]. During active flexion and abduction of the shoulder, the trapezius act as scapular retractor, and the serratus anterior enables the upward rotation and protraction of the scapulothoracic joint [3,32,33]. The rhomboids and levator scapulae mainly contribute to the scapula's retraction, elevation, and internal rotation [3,10]. In Figure 2a,b, a schematic representation of the main lines of action of the aforementioned muscles is presented.
The axillary fold is located below the glenohumeral joint connecting the humerus to the glenoid fossa of the scapula. In addition to being the site of a certain amount of fatty tissue and connective tissue, the axillary region posteriorly borders with the latissum dorsi muscle (Figure 2c) and teres major muscle (Figure 2d). The latissimus dorsi muscle is part of the muscles of the scapular movements enabling inferior angle pulling in multiple directions. Indeed, its multidirectional muscle fibers allow shoulder adduction, extension, and internal rotation. Besides allowing the movements (internal rotation and extension) of the humerus at the glenohumeral joint, the teres major muscle contributes to the scapular upward rotation and elevation [3]. The scapula serves as the location of various muscles' attachment [3]. Such muscles, having different sizes, functions, and depths, experience several stretching directions during upper limb movements in the different planes of the 3D space and at different degrees of elevation. Moreover, the scapula posteriorly is covered by overlying soft tissue, which in turn influences the superficial deformation of the scapular region. For all of these reasons, the skin deformations in the scapular region have dissimilarity in stretching position and magnitude during upper limbs motions. The main scapulothoracic muscles are the trapezius muscle, the serratus anterior muscle, the rhomboids, and the levator scapulae [3]. During active flexion and abduction of the shoulder, the trapezius act as scapular retractor, and the serratus anterior enables the upward rotation and protraction of the scapulothoracic joint [3,32,33]. The rhomboids and levator scapulae mainly contribute to the scapula's retraction, elevation, and internal rotation [3,10]. In Figure 2a,b, a schematic representation of the main lines of action of the aforementioned muscles is presented.
The axillary fold is located below the glenohumeral joint connecting the humerus to the glenoid fossa of the scapula. In addition to being the site of a certain amount of fatty tissue and connective tissue, the axillary region posteriorly borders with the latissum dorsi muscle (Figure 2c) and teres major muscle (Figure 2d). The latissimus dorsi muscle is part of the muscles of the scapular movements enabling inferior angle pulling in multiple directions. Indeed, its multidirectional muscle fibers allow shoulder adduction, extension, and internal rotation. Besides allowing the movements (internal rotation and extension) of the humerus at the glenohumeral joint, the teres major muscle contributes to the scapular upward rotation and elevation [3].

Participants
In this study, five male volunteers (mean ± standard deviation: age-25.4 ± 3.8 years old; body mass-74.8 ± 9.6 kg; height-1.77 ± 0.11 m; body mass index-23.7 ± 1.9 kg•m −2 ) with no history of shoulder pathologies were recruited. All participants performed the experimental tasks with their dominant (right) limb. Before experimental sessions, all subjects read and signed an informed consent, approved by the Ethical Committee of University Campus Bio-Medico of Rome (protocol code: 09/19 OSS ComEt UCBM). Table 1 shows the age and main anthropometric characteristics of the subjects involved in the study.

Experimental Set-Up
A Qualysis™ Motion Capture system (Qualysis AB, Gothenburg, Sweden) equipped with 10 Miqus M3 cameras (sampling frequency, 100 Hz) and 2 Miqus Video (sampling frequency, 25 Hz) was used to track a 6 × 5 grid of spherical retro-reflective markers (diameter, 8 mm). All markers were positioned on the right scapular region by the same investigator to avoid bias. Firstly, three markers were positioned on three skeletal landmarks of the scapula, i.e., angulus acromialis, trigonum spinae, and angulus inferior, identified by surface palpation. Then, the remaining 27 markers were positioned to form the 6 × 5 grid covering the entire scapular region of each subject (Figure 3a). Figure 3b,c show an actual reconstruction of the grid of markers during a task performed by a volunteer representing the starting position and elevated position, respectively.

Participants
In this study, five male volunteers (mean ± standard deviation: age-25.4 ± 3.8 years old; body mass-74.8 ± 9.6 kg; height-1.77 ± 0.11 m; body mass index-23.7 ± 1.9 kg·m −2 ) with no history of shoulder pathologies were recruited. All participants performed the experimental tasks with their dominant (right) limb. Before experimental sessions, all subjects read and signed an informed consent, approved by the Ethical Committee of University Campus Bio-Medico of Rome (protocol code: 09/19 OSS ComEt UCBM). Table 1 shows the age and main anthropometric characteristics of the subjects involved in the study.

Experimental Set-Up
A Qualysis™ Motion Capture system (Qualysis AB, Gothenburg, Sweden) equipped with 10 Miqus M3 cameras (sampling frequency, 100 Hz) and 2 Miqus Video (sampling frequency, 25 Hz) was used to track a 6 × 5 grid of spherical retro-reflective markers (diameter, 8 mm). All markers were positioned on the right scapular region by the same investigator to avoid bias. Firstly, three markers were positioned on three skeletal landmarks of the scapula, i.e., angulus acromialis, trigonum spinae, and angulus inferior, identified by surface palpation. Then, the remaining 27 markers were positioned to form the 6 × 5 grid covering the entire scapular region of each subject (Figure 3a). Figure 3b,c show an actual reconstruction of the grid of markers during a task performed by a volunteer representing the starting position and elevated position, respectively.

Experimental Protocol
Volunteers were verbally instructed by the same investigator, who also provided a practical demonstration of each task to be performed.
During experimental sessions, the starting position was with the arms along the body and palms towards the thighs. Figure 4 illustrates the movements investigated during experiments.

Experimental Protocol
Volunteers were verbally instructed by the same investigator, who also provided a practical demonstration of each task to be performed.
During experimental sessions, the starting position was with the arms along the body and palms towards the thighs. Figure 4 illustrates the movements investigated during experiments.

Experimental Protocol
Volunteers were verbally instructed by the same investigator, who also provided a practical demonstration of each task to be performed.
During experimental sessions, the starting position was with the arms along the body and palms towards the thighs. Figure 4 illustrates the movements investigated during experiments.  Task 6: 10 consecutive arm elevations in the scapular plane from starting position to maximum elevation.
All tasks were executed with the elbow fully extended and the thumb pointing upward. During each task, the same investigator guided the participants to perform the movements.

Motion Capture Data
The collected data were first pre-processed off-line using the Qualisys Track Manager (QTM) software (version 2021.1, Build 6300) for markers' labeling and trajectories gap filling by applying proprietary algorithms included in QTM software. All gap-filled trajectories were visually inspected. For further analysis, a process of manual identification of events corresponding to the starting and elevation positions reached by volunteers at each repetition was performed in QTM. Then, data of all subjects and executed tasks were exported to MATLAB (version 2020b). Markers' trajectories data were filtered using a low pass 4th order Butterworth filter with a cutoff frequency of 6 Hz. As there is no consensus on the directionality of deformation experienced in the scapular region during upper limb elevations, distances between pairs of markers were not calculated separately in the vertical and horizontal directions. Instead, distances between pairs of markers were computed by considering all possible combinations (i.e., 435) considering 30 elements (i.e., the number of markers) taken 2 at a time.
For each pair of markers, the distance D(i, j) between the i − th marker m(i) and the j − th marker m(j) was obtained as:

Skin Deformation Analysis and Statistics
For each pair of markers (i, j) with i = j, the skin relative strain variation ε k (i, j) at each k − th repetition was calculated using the following equation: where D(i, j) k and D(i, j) 0,k are the distances between the i − th and j − th markers corresponding to the k − th repetition at the elevated position and starting positions, respectively, and ∆D(i, j) k is the difference between the two mentioned distances. For greater clarity, Figure 5 shows the events corresponding to the starting position (light blue circle) and to the elevated position (green circle) for each repetition (in red). The mean percentage strain, ε%, was calculated as follows: A positive value of the mean strain ε% corresponds to the skin extension, while a negative value corresponds to the skin compression.
A positive value of the mean strain ̅ % corresponds to the skin extension, while a negative value corresponds to the skin compression. After calculating ̅ %, variations in skin strain were averaged among the five participants for each pair of markers. The descriptive analysis was performed by evaluating mean, median, standard deviation, minimum, and maximum strain. The Shapiro-Wilk test was used to assess the normality assumption of the data. If the Shapiro-Wilk test results were significant (p < 0.05), the nonparametric Wilcoxon rank-sum test was applied as a statistical method for strain comparison at 90° and maximum elevation in all planes. For all hypothesis tests, the p-value for significance was 0.05 for the rejection of the null hypothesis. Statistical analysis was performed in SPSS v28.0 (IBM, SPSS, Inc., Chicago, IL, USA).

Results
A total of 435 skin relative strain variations in the scapular region from 5 participants were analyzed during arm elevation in the frontal, sagittal, and scapular planes at 90° and maximum degree of elevation. During the elevation phase in all planes and at different degrees, some pairs of markers moved away, and others moved closer, suggesting extension and compression of the underlying scapular region, respectively. Figure 6 reports the distance trends of some pairs of markers during all tasks performed by a volunteer.
The Shapiro-Wilk test showed that strain distributions corresponding to different degrees of elevation were not normally distributed (Table 2). Moreover, the differences between strain at 90° and maximum elevation were significant, as shown by the results of the Wilcoxon rank-sum test (p < 0.05), see Table 2. Figure 7a reports the combination of box and violin plots to provide in a single representation the main features of strain distributions during the tasks performed in the frontal plane. The box plot allowed highlighting the mean value (represented by the asterisk), the median value (represented by the black horizontal line), and the interquartile After calculating ε%, variations in skin strain were averaged among the five participants for each pair of markers. The descriptive analysis was performed by evaluating mean, median, standard deviation, minimum, and maximum strain. The Shapiro-Wilk test was used to assess the normality assumption of the data. If the Shapiro-Wilk test results were significant (p < 0.05), the nonparametric Wilcoxon rank-sum test was applied as a statistical method for strain comparison at 90 • and maximum elevation in all planes. For all hypothesis tests, the p-value for significance was 0.05 for the rejection of the null hypothesis. Statistical analysis was performed in SPSS v28.0 (IBM, SPSS, Inc., Chicago, IL, USA).

Results
A total of 435 skin relative strain variations in the scapular region from 5 participants were analyzed during arm elevation in the frontal, sagittal, and scapular planes at 90 • and maximum degree of elevation. During the elevation phase in all planes and at different degrees, some pairs of markers moved away, and others moved closer, suggesting extension and compression of the underlying scapular region, respectively. Figure 6 reports the distance trends of some pairs of markers during all tasks performed by a volunteer.
The Shapiro-Wilk test showed that strain distributions corresponding to different degrees of elevation were not normally distributed (Table 2). Moreover, the differences between strain at 90 • and maximum elevation were significant, as shown by the results of the Wilcoxon rank-sum test (p < 0.05), see Table 2.   From the analysis of Figure 7a is clear the greater dispersion of the ̅ % during the task at maximum elevation (in blue) than task up to about 90° (in yellow). For maximum elevation in the frontal plane, results of ̅ % showed a mean ± standard deviation equals to −0.36 ± 13.27, a median of −4.05, and an IQR of −8.10-3.58. For 90° of elevation in the frontal plane, the mean ± standard deviation was −0.46 ± 6.43, the median was −0.80, and the IQR was −3.39-1.86. The bigger extension of the IQR calculated during maximum extension confirms the higher dispersion in this task.
A similar analysis has been performed considering the ̅ % absolute values reported in Figure 7b. Such analysis allows comparing the ̅ % experienced during the two degrees of elevation by focusing on the skin strain's amplitude without discriminating between compression and extension. From the analysis of Figure 7b, it is clear that during the task  From the analysis of Figure 7a is clear the greater dispersion of the ε% during the task at maximum elevation (in blue) than task up to about 90 • (in yellow). For maximum elevation in the frontal plane, results of ε% showed a mean ± standard deviation equals to −0.36 ± 13.27, a median of −4.05, and an IQR of −8.10-3.58. For 90 • of elevation in the frontal plane, the mean ± standard deviation was −0.46 ± 6.43, the median was −0.80, and the IQR was −3.39-1.86. The bigger extension of the IQR calculated during maximum extension confirms the higher dispersion in this task.
A similar analysis has been performed considering the ε% absolute values reported in Figure 7b. Such analysis allows comparing the ε% experienced during the two degrees of elevation by focusing on the skin strain's amplitude without discriminating between compression and extension. From the analysis of Figure 7b, it is clear that during the task at maximum elevation (in orange), the absolute value of ε% is bigger than the one up to about 90 • (in green). For maximum elevation in the frontal plane, the mean ± standard deviation was 9.90 ± 8.84, the median was 7.30, and the IQR was 4.02-12.56. For 90 • , the mean ± standard deviation was 4.46 ± 4.64, the median was 2.60, and the IQR was 1.21−6.23. These results highlight that skin strains in the scapular region are greatest during maximal abduction and are also confirmed by the maximum ε% value (i.e., 52.95% for maximum elevation vs. 28.26% for elevation up to 90 • ). The region that underwent maximum extension corresponds to the pair of markers 19-20 for both degrees of elevation in the frontal plane (Figure 7c,d).
at maximum elevation (in orange), the absolute value of ̅ % is bigger than the one up to about 90° (in green). For maximum elevation in the frontal plane, the mean ± standard deviation was 9.90 ± 8.84, the median was 7.30, and the IQR was 4.02-12.56. For 90°, the mean ± standard deviation was 4.46 ± 4.64, the median was 2.60, and the IQR was 1.21−6.23. These results highlight that skin strains in the scapular region are greatest during maximal abduction and are also confirmed by the maximum ̅ % value (i.e., 52.95% for maximum elevation vs. 28.26% for elevation up to 90°). The region that underwent maximum extension corresponds to the pair of markers 19-20 for both degrees of elevation in the frontal plane (Figure 7c,d). The region that underwent maximum compression during upper arm abduction corresponds to the first line of the grid for both degrees of elevation (Figure 7c,d). Table 3 reports the extreme values of ̅ % for both extension and compression during tasks performed in the frontal plane. Data confirm that skin strains are bigger during maximal abduction.
From the analysis of Figure 8a is clear the greater dispersion of the ̅ % during the task at maximum elevation (in blue) than task up to about 90° (in yellow) performed in the sagittal plane. For maximum elevation in the sagittal plane, results of ̅ % showed a mean ± standard deviation equals to −6.87 ± 14.62, a median of −3.96, and an IQR of −9.68-7.50. For 90°, the mean ± standard deviation was −5.37 ± 11.59, the median was −2.55, and the IQR was −3.77-12.56. Also in this case, from the analysis of Figure 8b is clear that during the task at maximum elevation (in orange) the absolute value of % is bigger than the one up to about 90° (in green). For maximum elevation in the sagittal plane, the mean ± standard deviation was 11.30 ± 9.28, the median was 9.10, and the IQR was 5.32-14.23. For 90°, the mean ± standard deviation was 9.40 ± 8.64, the median was 6.47, and the IQR was 3.39-12.86. The region that underwent maximum compression during upper arm abduction corresponds to the first line of the grid for both degrees of elevation (Figure 7c,d). Table 3 reports the extreme values of ε% for both extension and compression during tasks performed in the frontal plane. Data confirm that skin strains are bigger during maximal abduction.
From the analysis of Figure 8a is clear the greater dispersion of the ε% during the task at maximum elevation (in blue) than task up to about 90 • (in yellow) performed in the sagittal plane. For maximum elevation in the sagittal plane, results of ε% showed a mean ± standard deviation equals to −6.87 ± 14.62, a median of −3.96, and an IQR of −9.68-7.50. For 90 • , the mean ± standard deviation was −5.37 ± 11.59, the median was −2.55, and the IQR was −3.77-12.56.
Also in this case, from the analysis of Figure 8b is clear that during the task at maximum elevation (in orange) the absolute value of ε% is bigger than the one up to about 90 • (in green). For maximum elevation in the sagittal plane, the mean ± standard deviation was 11.30 ± 9.28, the median was 9.10, and the IQR was 5.32-14.23. For 90 • , the mean ± standard deviation was 9.40 ± 8.64, the median was 6.47, and the IQR was 3.39-12.86.
The maximum positive values were found to be 60.87% and 60.12% for maximum and 90 • of elevation, respectively ( Table 4). The region that underwent maximum extension corresponds to the pair of markers 19-20 for both degrees of elevation in the sagittal plane (Figure 8c,d). Unlike movements performed in the frontal plane, during upper arm elevations in the sagittal plane, the pairs of markers corresponding to the maximum compressive strain values were not distributed along the same direction (Figure 8c,d). Table 4 reports the extreme values of ε% for both extension and during tasks performed in the sagittal plane.  The maximum positive values were found to be 60.87% and 60.12% for maximum and 90° of elevation, respectively ( Table 4). The region that underwent maximum extension corresponds to the pair of markers 19-20 for both degrees of elevation in the sagittal plane (Figure 8c,d). Unlike movements performed in the frontal plane, during upper arm elevations in the sagittal plane, the pairs of markers corresponding to the maximum compressive strain values were not distributed along the same direction (Figure 8c,d). Table 4 reports the extreme values of ̅ % for both extension and during tasks performed in the sagittal plane.   Figure 9a shows strain distributions of the scapular region corresponding to the elevations performed in the scapular plane. For maximum elevation in the scapular plane, results of ε% showed a mean ± standard deviation equals to 0.32 ± 11.08, a median of −2.96, and an IQR of −6.56-4.39. For 90 • of elevation in the scapular plane, the mean ± standard deviation was 1.86 ± 7.94, the median was −0.60, and the IQR was −3.22-5.57.   Figure 9a shows strain distributions of the scapular region corresponding to the elevations performed in the scapular plane. For maximum elevation in the scapular plane, results of ̅ % showed a mean ± standard deviation equals to 0.32 ± 11.08, a median of −2.96, and an IQR of −6.56-4.39. For 90° of elevation in the scapular plane, the mean ± standard deviation was 1.86 ± 7.94, the median was −0.60, and the IQR was −3.22-5.57. As in the two previous planes, from the analysis of Figure 9b is clear that during the task at maximum elevation (in orange) in the scapular plane, the absolute value of ̅ % is bigger than the one up to about 90° (in green). For maximum elevation in the scapular As in the two previous planes, from the analysis of Figure 9b is clear that during the task at maximum elevation (in orange) in the scapular plane, the absolute value of ε% is bigger than the one up to about 90 • (in green). For maximum elevation in the scapular plane, the mean ± standard deviation was 8.28 ± 7.36, the median was 6.05, and the IQR was 3.34-10.70. For 90 • , the mean ± standard deviation was 5.79 ± 5.74, the median was 3.87, and the IQR was 1.91-7.76.
The maximum positive values were 48.20% and 40.89% for maximum and 90 • of elevation, respectively ( Table 5). The region that underwent maximum extension corresponds to the pair of markers 19-20 for both degrees of elevation in the scapular plane (Figure 9c,d). As in the case of movements performed in the sagittal plane, also during the elevation of the upper limb in the scapular plane, the distribution of the pairs of markers corresponding to the maximum compressive strain values is not concentrated along the same row of the grid of markers. Even in this case, the region that underwent greater extension was the one surrounding the axillary fold, although along slightly different directions than in the other planes (Figure 9c,d). Table 5 reports the extreme values of ε% for both extension and compression during tasks performed in the scapular plane.

Discussion
Monitoring scapular movements may be useful in rehabilitation and clinical research. This study proposes a methodological approach to quantify scapular skin strain using a 6 × 5 grid of retro-reflective markers. We implemented this method for upper limb flexion in the sagittal plane, elevation in the scapular plane (scaption), and abduction in the frontal plane. This analysis may be fundamental for the development of some solutions able to monitor the scapular movements. Indeed, an open challenge in the development of wearable systems based on strain sensors is the proper placement of the sensing elements. To date, several textile-based strain sensors have been designed and employed to measure human joints movements [7,8,22,[34][35][36][37]. Among textile-based strain sensors, resistive ones are popular for instrumenting wearables [6,19]. These sensors are mainly composed of an elastic textile substrate and conductive materials, which undergo microstructural changes in response to an applied deformation resulting in electrical resistance variation in the sensing elements [6,7]. The textile component enables the integration into garments as adherent as possible or into polymeric substrates that could potentially be directly applied to the skin. The textile component allows the sensitive element to stretch and relax during movements, thanks to its elastic characteristics. One of the main requirements for developing wearable systems integrating textile-based strain sensors is that they should adhere perfectly to the surface of the body region of interest. Moreover, improper orientations of the sensors could negatively influence the sensitivity for joints movements detection. As regard scapular movements, the unreliable reading of textile-based strain sensors is further influenced by the simultaneity of translations and rotations that the scapula undergoes during upper limb movements. For this reason, identifying the areas in the scapular region that experience the greatest deformation could provide useful information about the design, integration, and placement of textile-based wearable strain sensors. In a previous study [30], skin strain field analysis in the region surrounding the shoulder joint was performed using three-dimensional image correlation technique. Shoulder abduction and flexion were investigated in a single volunteer, showing that the area that experienced more significant strains corresponds to that surrounding the axillary fold posteriorly, in accordance with our findings. Unlike our study, in [13], a grid of markers was placed on the scapular region to obtain a surface mapping from which to infer the scapular kinematics.
In the present study, the motion tracking data were used to provide the distribution of length changes in the posterior scapular region, calculated in terms of distance between all possible combinations of markers pairs. Strain distribution (Figures 7-9) shows interesting characteristics for all movements performed in all planes and degrees of elevation. Namely, the region with the highest extension was the area surrounding the axillary fold. Although this region corresponds to an area with a greater amount of underlying soft tissue, it also has a high number of muscles, which contract during arms elevation, inducing a corresponding surface deformation. Results showed a significant difference between elevation up to 90 • and maximum elevation for all the performed tasks. Concerning the positive strains (i.e., extension), the highest percentage positive strain was found to be: 28.26% and 52.95% for elevation in the frontal plane up to 90 • and maximum elevation, respectively; 60.12% and 60.87% for elevation in the sagittal plane up to 90 • and maximum elevation, respectively; and 40.89% and 48.20% for elevation in the scapular plane up to 90 • and maximum elevation, respectively. In all these cases, the maximum extension is referred to the pair of markers 19-20 placed horizontally near the axillary fold (see Figure 3). Conversely, the same generality of results cannot be applied to regions that underwent maximum compression. Although the regions subjected to the greatest compression mostly correspond to the first rows of the marker grid (Figure 7c,d, Figures 8c and 9c,d), in some cases, the pairs of markers that experienced the greatest compression are arranged in different regions (Figure 8c,d and Figure 9c). The reason for these results is probably related to the anthropometric heterogeneity of the subjects involved in the experimental trials. This aspect is not of particular relevance in the design of wearable systems based on resistive textile sensors since they work better in extension than in shortening [34]. Therefore, the regions subjected to higher stretch values should be considered for the placement of textile-based strain sensors.
The absence of deep analysis on the skin strain in the scapular region is highlighted by the different positioning and number of resistive textile-based strain sensors used in wearable systems designed for monitoring scapular movements [7,8,22]. Although these studies showed promising results about monitoring scapular movements in healthy subjects and patients with musculoskeletal or neurological disorders, they all empirically placed the sensors on the scapular region.

Conclusions
In conclusion, this study proposed a new method for skin strain analysis of the scapular region. The method was used to estimate the skin scapular surface strain on five volunteers during upper limb movements of clinical relevance. This is the first study investigating skin deformation of the scapular region induced by arms elevation in different planes and at different degrees of elevation. The results suggested interesting insights for the integration and positioning of resistive textile-based strain sensors within wearable systems for monitoring scapular movements.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.