Fully Automated Analysis of Muscle Architecture from B-Mode Ultrasound Images with DL_Track_US.

OBJECTIVE: B-mode ultrasound can be used to image musculoskeletal tissues, but one major bottleneck is analyses of muscle architectural parameters (i.e., muscle thickness, pennation angle and fascicle length), which are most often performed manually. METHODS: In this study we trained two different neural networks (classic U-Net and U-Net with VGG16 pre-trained encoder) to detect muscle fascicles and aponeuroses using a set of labeled musculoskeletal ultrasound images. We determined the best-performing model based on intersection over union and loss metrics. We then compared neural network predictions on an unseen test set with those obtained via manual analysis and two existing semi/automated analysis approaches (simple muscle architecture analysis [SMA] and UltraTrack). DL_Track_US detects the locations of the superficial and deep aponeuroses, as well as multiple fascicle fragments per image. RESULTS: For single images, DL_Track_US yielded results similar to those produced by a non-trainable automated method (SMA; mean difference in fascicle length: 5.1 mm) and human manual analysis (mean difference: -2.4 mm). Between-method differences in pennation angle were within 1.5°, and mean differences in muscle thickness were less than 1 mm. Similarly, for videos, there was overlap between the results produced with UltraTrack and DL_Track_US, with intraclass correlations ranging between 0.19 and 0.88. CONCLUSION: DL_Track_US is fully automated and open source and can estimate fascicle length, pennation angle and muscle thickness from single images or videos, as well as from multiple superficial muscles. We also provide a user interface and all necessary code and training data for custom model development.

Medial gastrocnemius muscle fascicle function during heel-rise after non-operative repair of Achilles tendon rupture.

BACKGROUND: To better understand muscle remodelling in dynamic conditions after an Achilles tendon rupture, this study examined the length of medial gastrocnemius muscle fascicles during a heel-rise at 6- and 12-months after non-operative ATR treatment. METHODS: Participants (15 M, 3F) were diagnosed with acute Achilles tendon rupture. Medial gastrocnemius subtendon length, fascicle length and pennation angle were assessed in resting conditions, and fascicle shortening during bi- and unilateral heel-rises. FINDINGS: Fascicle shortening was smaller on the injured side (mean difference [95% CI]: -9.7 mm [-14.7 to -4.7 mm]; -11.1 mm [-16.5 to -5.8 mm]) and increased from 6- to 12 months (4.5 mm [2.8-6.3 mm]; 3.2 mm [1.4-4.9 mm]) in bi- and unilateral heel-rise, respectively. The injured tendon was longer compared to contralateral limb (2.16 cm [0.54-3.79 cm]) and the length decreased over time (-0.78 cm [-1.28 to -0.29 cm]). Tendon length correlated with fascicle shortening in bilateral (r = -0.671, p = 0.002; r = -0.666, p = 0.003) and unilateral (r = -0.773, p ≤ 0.001; r = -0.616, p = 0.006) heel-rise, at 6- and 12-months, respectively. In the injured limb, the change over time in fascicle shortening correlated with change in subtendon length in unilateral heel-rise (r = 0.544, p = 0.02). INTERPRETATION: This study showed that the lengths of the injured tendon and associated muscle can adapt throughout the first year after rupture when patients continue physiotherapy and physical exercises. For muscle, measures of resting length may not be very informative about adaptations, which manifest themselves during functional tasks such as unilateral heel-rise.

Exploration of muscle-tendon biomechanics one year after Achilles tendon rupture and the compensatory role of flexor hallucis longus.

Achilles tendon (AT) rupture leads to long-term structural and functional impairments. Currently, the predictors of good recovery after rupture are poorly known. Thus, we aimed to explore the interconnections between structural, mechanical, and neuromuscular parameters and their associations with factors that could explain good recovery in patients with non-surgically treated AT rupture. A total of 35 patients with unilateral rupture (6 females) participated in this study. Muscle-tendon structural, mechanical, and neuromuscular parameters were measured 1-year after rupture. Interconnections between the inter-limb differences (Δ) were explored using partial correlations, followed by multivariable linear regression to find associations between the measured factors and the following markers that indicate good recovery: 1) tendon length, 2) tendon non-uniform displacement, and 3) flexor hallucis longus (FHL) normalized EMG amplitude difference between limbs. Δmedial gastrocnemius (MG) (ß = -0.12, p = 0.007) and Δlateral gastrocnemius (ß = -0.086, p = 0.030) subtendon lengths were associated with MG tendon Δstiffness. MG (ß = 11.56, p = 0.003) and soleus (ß = 2.18, p = 0.040) Δsubtendon lengths explained 48 % of variance in FHL EMG amplitude. Regression models for tendon length and non-uniform displacement were not significant. Smaller inter-limb differences in Achilles subtendon lengths were associated with smaller differences in the AT stiffness between limbs, and a smaller contribution of FHL muscle to the plantarflexion torque. In the injured limb, the increased contribution of FHL appears to partially counteract a smaller contribution from MG due to the elongated tendon, however the role of FHL should not be emphasized during rehabilitation to allow recovery of the TS muscles.

Accuracy of a computer vision system for estimating biomechanical measures of body function in axial spondyloarthropathy patients and healthy subjects.

OBJECTIVE: Advances in computer vision make it possible to combine low-cost cameras with algorithms, enabling biomechanical measures of body function and rehabilitation programs to be performed anywhere. We evaluated a computer vision system's accuracy and concurrent validity for estimating clinically relevant biomechanical measures. DESIGN: Cross-sectional study. SETTING: Laboratory. PARTICIPANTS: Thirty-one healthy participants and 31 patients with axial spondyloarthropathy. INTERVENTION: A series of clinical functional tests (including the gold standard Bath Ankylosing Spondylitis Metrology Index tests). Each test was performed twice: the first performance was recorded with a camera, and a computer vision algorithm was used to estimate variables. During the second performance, a clinician measured the same variables manually. MAIN MEASURES: Joint angles and inter-limb distances. Clinician measures were compared with computer vision estimates. RESULTS: For all tests, clinician and computer vision estimates were correlated (r2 values: 0.360-0.768). There were no significant mean differences between methods for shoulder flexion (left: 2 ± 14° (mean ± standard deviation), t = 0.99, p < 0.33; right: 3 ± 15°, t = 1.57, p < 0.12), side flexion (left: - 0.5 ± 3.1 cm, t = -1.34, p = 0.19; right: 0.5 ± 3.4 cm, t = 1.05, p = 0.30) and lumbar flexion ( - 1.1 ± 8.2 cm, t = -1.05, p = 0.30). For all other movements, significant differences were observed, but could be corrected using a systematic offset. CONCLUSION: We present a computer vision approach that estimates distances and angles from clinical movements recorded with a phone or webcam. In the future, this approach could be used to monitor functional capacity and support physical therapy management remotely.

Understanding altered contractile properties in advanced age: insights from a systematic muscle modelling approach.

Age-related alterations of skeletal muscle are numerous and present inconsistently, and the effect of their interaction on contractile performance can be nonintuitive. Hill-type muscle models predict muscle force according to well-characterised contractile phenomena. Coupled with simple, yet reasonably realistic activation dynamics, such models consist of parameters that are meaningfully linked to fundamental aspects of muscle excitation and contraction. We aimed to illustrate the utility of a muscle model for elucidating relevant mechanisms and predicting changes in output by simulating the individual and combined effects on isometric force of several known ageing-related adaptations. Simulating literature-informed reductions in free Ca2+ concentration and Ca2+ sensitivity generated predictions at odds qualitatively with the characteristic slowing of contraction speed. Conversely, incorporating slower Ca2+ removal or a fractional increase in type I fibre area emulated expected changes; the former was required to simulate slowing of the twitch measured experimentally. Slower Ca2+ removal more than compensated for force loss arising from a large reduction in Ca2+ sensitivity or moderate reduction in Ca2+ release, producing realistic age-related shifts in the force-frequency relationship. Consistent with empirical data, reductions in free Ca2+ concentration and Ca2+ sensitivity reduced maximum tetanic force only slightly, even when acting in concert, suggesting a modest contribution to lower specific force. Lower tendon stiffness and slower intrinsic shortening speed slowed and prolonged force development in a compliance-dependent manner without affecting force decay. This work demonstrates the advantages of muscle modelling for exploring sources of variation and identifying mechanisms underpinning the altered contractile properties of aged muscle.

Feasibility of OpenPose markerless motion analysis in a real athletics competition.

This study tested the performance of OpenPose on footage collected by two cameras at 200 Hz from a real-life competitive setting by comparing it with manually analyzed data in SIMI motion. The same take-off recording from the men's Long Jump finals at the 2017 World Athletics Championships was used for both approaches (markerless and manual) to reconstruct the 3D coordinates from each of the camera's 2D coordinates. Joint angle and Centre of Mass (COM) variables during the final step and take-off phase of the jump were determined. Coefficients of Multiple Determinations (CMD) for joint angle waveforms showed large variation between athletes with the knee angle values typically being higher (take-off leg: 0.727 ± 0.242; swing leg: 0.729 ± 0.190) than those for hip (take-off leg: 0.388 ± 0.193; swing leg: 0.370 ± 0.227) and ankle angle (take-off leg: 0.247 ± 0.172; swing leg: 0.155 ± 0.228). COM data also showed considerable variation between athletes and parameters, with position (0.600 ± 0.322) and projection angle (0.658 ± 0.273) waveforms generally showing better agreement than COM velocity (0.217 ± 0.241). Agreement for discrete data was generally poor with high random error for joint kinematics and COM parameters at take-off and an average ICC across variables of 0.17. The poor agreement statistics and a range of unrealistic values returned by the pose estimation underline that OpenPose is not suitable for in-competition performance analysis in events such as the long jump, something that manual analysis still achieves with high levels of accuracy and reliability.

DeepACSA: Automatic Segmentation of Cross-Sectional Area in Ultrasound Images of Lower Limb Muscles Using Deep Learning.

PURPOSE: Muscle anatomical cross-sectional area (ACSA) can be assessed using ultrasound and images are usually evaluated manually. Here, we present DeepACSA, a deep learning approach to automatically segment ACSA in panoramic ultrasound images of the human rectus femoris (RF), vastus lateralis (VL), gastrocnemius medialis (GM) and lateralis (GL) muscles. METHODS: We trained three muscle-specific convolutional neural networks (CNN) using 1772 ultrasound images from 153 participants (age = 38.2 yr, range = 13-78). Images were acquired in 10% increments from 30% to 70% of femur length for RF and VL and at 30% and 50% of muscle length for GM and GL. During training, CNN performance was evaluated using intersection-over-union scores. We compared the performance of DeepACSA to manual analysis and a semiautomated algorithm using an unseen test set. RESULTS: Comparing DeepACSA analysis of the RF to manual analysis with erroneous predictions removed (3.3%) resulted in intraclass correlation (ICC) of 0.989 (95% confidence interval = 0.983-0.992), mean difference of 0.20 cm 2 (0.10-0.30), and SEM of 0.33 cm 2 (0.26-0.41). For the VL, ICC was 0.97 (0.96-0.968), mean difference was 0.85 cm 2 (-0.4 to 1.31), and SEM was 0.92 cm 2 (0.73-1.09) after removal of erroneous predictions (7.7%). After removal of erroneous predictions (12.3%), GM/GL muscles demonstrated an ICC of 0.98 (0.96-0.99), a mean difference of 0.43 cm 2 (0.21-0.65), and an SEM of 0.41 cm 2 (0.29-0.51). Analysis duration was 4.0 ± 0.43 s (mean ± SD) for analysis of one image in our test set using DeepACSA. CONCLUSIONS: DeepACSA provides fast and objective segmentation of lower limb panoramic ultrasound images comparable with manual segmentation. Inaccurate model predictions occurred predominantly on low-quality images, highlighting the importance of high-quality image for accurate prediction.

In vivo localized gastrocnemius subtendon representation within the healthy and ruptured human Achilles tendon.

The Achilles tendon (AT) is composed of three distinct in-series elastic subtendons, arising from different muscles in the triceps surae. Independent activation of any of these muscles is thought to induce sliding between the adjacent AT subtendons. We aimed to investigate displacement patterns during voluntary contraction (VOL) and selective transcutaneous stimulation of medial (MGstim) and lateral (LGstim) gastrocnemius between ruptured and healthy tendons and to examine the representative areas of AT subtendons. Twenty-eight patients with unilateral AT rupture performed bilateral VOL at 30% of the maximal isometric uninjured plantarflexion torque. AT displacement was analyzed from sagittal B-mode ultrasonography images during VOL, MGstim, and LGstim. Three-way ANOVA revealed a significant two-way interaction of contraction type × location on the tendon displacement [F(10-815) = 3.72, P < 0.001]. The subsequent two-way analysis revealed a significant contraction type × location interaction for tendon displacement [F(10-410)=3.79, P < 0.001] in the uninjured limb only, where LGstim displacement pattern was significantly different from MGstim (P = 0.008) and VOL (P = 0.005). When comparing contraction types between limbs, there were no difference in the displacement patterns, but displacement amplitudes differed. There was no significant difference in the location of maximum or minimum displacement between limbs. The displacement pattern was not different in nonsurgically treated compared with uninjured tendons 1 yr after rupture. However, free tendon stiffness seems to be lower in the injured AT, leading to more displacement during electrically induced contractions compared with the uninjured. Our results suggest that near the calcaneus, LG subtendon is located in the most anterior region adjacent to medial gastrocnemius.NEW & NOTEWORTHY Using selective electrical stimulation, we report the distributions of medial and lateral gastrocnemius subtendon representations within the healthy and ruptured Achilles tendon. In the majority of our sample, lateral gastrocnemius subtendon was found in the most anterior region adjacent to medial gastrocnemius both in the healthy and ruptured, nonsurgically treated tendon. The tendon internal displacement pattern does not seem to differ, but displacement amplitude and nonuniformity differed between healthy and ruptured tendons 1 yr after rupture.

Tendon length estimates are influenced by tracking location.

PURPOSE: Measurement of medial gastrocnemius (MG) tendon length using ultrasonography (US) requires the muscle-tendon junction (MTJ) to be located. Previously, the MG MTJ has been tracked from different proximo-distal locations near the MTJ, which could influence estimates of tendon length change due to the different characteristics of the aponeurosis and tendon. We used US to evaluate the effect of tracking point location on MG MTJ displacement during maximal and submaximal (10, 20 and 30% of the non-injured maximal) isometric plantar flexion contractions. METHODS: Displacement behaviour of MTJ was tracked from (1) the exact MTJ; and (2) from an insertion point of a muscle fascicle on the aponeurosis 1.3 ± 0.6 cm proximal to the MTJ, in both limbs of patients with unilateral Achilles tendon rupture (ATR) (n = 22, 4 females, 42 ± 9 years, 177 ± 9 cm, 79 ± 10 kg). RESULTS: In the non-injured limb, displacement (1.3 ± 0.5 cm vs. 1.1 ± 0.6 cm) and strain (6.7 ± 2.8% vs. 5.8 ± 3.3%) during maximal voluntary contraction were larger when tracking a point on the aponeurosis than when tracking the MTJ (both p < 0.001). The same was true for all contraction levels, and both limbs. CONCLUSION: Tracking a point on the aponeurosis consistently exaggerates estimates of tendon displacement, and the magnitude of this effect is contraction intensity-dependent. When quantifying displacement and strain of the Achilles tendon, the MTJ should be tracked directly, rather than tracking a surrogate point proximal to the MTJ. The latter method includes part of the aponeurosis, which due to its relative compliance, artificially increases estimates of MTJ displacement and strain.

Muscle-tendon morphomechanical properties of non-surgically treated Achilles tendon 1-year post-rupture.

BACKGROUND: Achilles tendon rupture appears to alter stiffness and length of the tendon. These alterations may affect the function of tendon in force transmission and in energy storage and recovery. We studied the mechanical properties of the Achilles' tendon post-rupture and their association with function. METHODS: Twenty-four (20 males, 4 females) participants (mean age: 43 y, 176 cm, 81 kg) were recruited. Ultrasonography and dynamometry were used to assess the muscle-tendon unit morphological and mechanical properties of non-surgically treated patients 1-year post rupture. FINDINGS: Injured tendons were longer with difference of 1.8 cm (95%CI: 0.5-1.9 cm; P < 0.001), and thicker by 0.2 mm (0.2-0.3 mm; P < 0.01). Medial gastrocnemius cross-sectional area was 1.0 cm2 smaller (0.8-1.1 cm2; P < 0.001), fascicles were 0.6 cm shorter (0.5-0.7 cm; P < 0.001) and pennation angle was 2.5° higher (1.3-3.6°; P < 0.001) when compared to the uninjured limb. We found no differences between injured and uninjured tendon stiffness 1-year post-rupture (mean difference: 29.8 N/mm, -7.7-67.3 N/mm; P = 0.170). The injured tendon showed 1.8 mm (1.2-2.4 mm; P < 0.01) lower elongation during maximal voluntary isometric contractions. Patient-reported functional outcome was related to the tendon resting length (ß = 0.68, r(10) = 4.079, P = 0.002). Inter-limb differences in the medial gastrocnemius fascicle length were related to inter-limb differences in maximum contractions (ß = 1.17, r(14) = 2.808, P = 0.014). INTERPRETATION: Longer Achilles tendon resting length was associated with poorer self-evaluated functional outcome. Although the stiffness of non-surgically treated and uninjured tendons was similar 1-year post rupture, plantar flexion strength deficit was still present, possibly due to shorter medial gastrocnemius fascicle length.

Repeatability and sensitivity of passive mechanical stiffness measurements in the triceps surae muscle-tendon complex.

Measurements of muscle-tendon unit passive mechanical properties are often used to illustrate acute and chronic responses to a training stimulus. The purpose of this study was to quantify the inter-session repeatability of triceps surae passive stiffness measurements in athletic and non-athletic populations, with the view to discussing its usefulness both as a muscle-tendon profiling tool and a control measure for studies with multiple data collection sessions. The study also aimed to observe the effects of quiet standing on passive stiffness parameters. Twenty-nine men (10 cyclists, nine triathletes, 10 controls) visited the laboratory on three separate occasions, where passive stiffness tests were carried out using an isokinetic dynamometer and B-mode ultrasound. Participants were fully rested on two of the sessions and subjected to 20 min of quiet standing in the other. The passive stiffness assessment generally showed only moderate inter-session repeatability but was still able to detect inter-group differences, with triathletes showing higher passive stiffness than cyclists (p < 0.05). Furthermore, quiet standing impacted passive stiffness by causing a reduction in ankle joint range of motion, although mechanical resistance to stretch in the muscle-tendon unit at a given joint angle was relatively unaffected. These findings show that passive stiffness assessment is appropriate for detecting inter-group differences in the triceps surae and even the effects of a low-intensity task such as quiet standing, despite showing some inter-session variation. However, the inter-session variation suggests that passive stiffness testing might not be suitable as a control measure when testing participants on multiple sessions.

Using deep neural networks for kinematic analysis: Challenges and opportunities.

Kinematic analysis is often performed in a lab using optical cameras combined with reflective markers. With the advent of artificial intelligence techniques such as deep neural networks, it is now possible to perform such analyses without markers, making outdoor applications feasible. In this paper I summarise 2D markerless approaches for estimating joint angles, highlighting their strengths and limitations. In computer science, so-called "pose estimation" algorithms have existed for many years. These methods involve training a neural network to detect features (e.g. anatomical landmarks) using a process called supervised learning, which requires "training" images to be manually annotated. Manual labelling has several limitations, including labeller subjectivity, the requirement for anatomical knowledge, and issues related to training data quality and quantity. Neural networks typically require thousands of training examples before they can make accurate predictions, so training datasets are usually labelled by multiple people, each of whom has their own biases, which ultimately affects neural network performance. A recent approach, called transfer learning, involves modifying a model trained to perform a certain task so that it retains some learned features and is then re-trained to perform a new task. This can drastically reduce the required number of training images. Although development is ongoing, existing markerless systems may already be accurate enough for some applications, e.g. coaching or rehabilitation. Accuracy may be further improved by leveraging novel approaches and incorporating realistic physiological constraints, ultimately resulting in low-cost markerless systems that could be deployed both in and outside of the lab.

Non-uniform displacement within ruptured Achilles tendon during isometric contraction.

The purpose of this study was investigate tendon displacement patterns in non-surgically treated patients 14 months after acute Achilles tendon rupture (ATR) and to classify patients into groups based on their Achilles tendon (AT) displacement patterns. Twenty patients were tested. Sagittal images of AT were acquired using B-mode ultrasonography during ramp contractions at a torque level corresponding to 30% of the maximal isometric plantarflexion torque of the uninjured limb. A speckle tracking algorithm was used to track proximal-distal movement of the tendon tissue at 6 antero-posterior locations. Two-way repeated measures ANOVA for peak tendon displacement was performed. K-means clustering was used to classify patients according to AT displacement patterns. The difference in peak relative displacement across locations was larger in the uninjured (1.29 ± 0.87 mm) than the injured limb (0.69 ± 0.68 mm), with a mean difference (95% CI) of 0.60 mm (0.14-1.05 mm, P < .001) between limbs. For the uninjured limb, cluster analysis formed 3 groups, while 2 groups were formed for the injured limb. The three distinct patterns of AT displacement during isometric plantarflexion in the uninjured limb may arise from subject-specific anatomical variations of AT sub-tendons, while the two patterns in the injured limb may reflect differential recovery after ATR with non-surgical treatment. Subject-specific tendon characteristics are a vital determinant of stress distribution across the tendon. Changes in stress distribution may lead to variation in the location and magnitude of peak displacement within the free AT. Quantifying internal tendon displacement patterns after ATR provides new insights into AT recovery.

Validity of long-term and short-term recall of occupational sitting time in Finnish and Chinese office workers.

BACKGROUND: As sedentary behavior is a global health issue, there is a need for methods of self-reported sitting assessment. The accuracy and reliability of these methods should also be tested in various populations and different cultural contexts. This study examined the validity of long-term and short-term recall of occupational sitting time in Finnish and Chinese subgroups. METHODS: Two cohort groups of office-based workers (58.6% female, age range 22-67 years) participated: a Finnish group (FIN, n = 34) and a Chinese group (CHI, n = 36). Long-term (past 3-month sitting) and short-term (daily sitting assessed on 5 consecutive days) single-item measures were used to assess self-reported occupational sitting time. Values from each participant were compared to objectively measured occupational sitting time assessed via thigh-mounted accelerometers, with Spearman's rho (ρ) used to assess validity and the Bland-Altman method used to evaluate agreement. Coefficients of variation depicted day-to-day variability of time spent on sitting at work. RESULTS: In the total study sample, the results showed that both long-term and short-term recall correlated with accelerometer-derived sitting time (ρ = 0.532, 95% confidence intervals (CI): 0.336-0.684, p< 0.001; ρ = 0.533, 95%CI: 0.449-0.607, p< 0.001, respectively). Compared to objectively measured sitting time, self-reported occupational sitting time was 2.4% (95%CI: -0.5% to 5.3%, p = 0.091) and 2.2% (95%CI: 0.7%-3.6%, p = 0.005) greater for long-term and short-term recall, respectively. The agreement level was within the range -21.2% to 25.9% for long-term recall, and -24.2% to 28.5% for short-term recall. During a 5-day work week, day-to-day variation of sitting time was 9.4% ± 11.4% according to short-term recall and 10.4% ± 8.4% according to accelerometry-derived occupational sitting time. CONCLUSION: Overall, both long-term and short-term self-reported instruments provide acceptable measures of occupational sitting time in an office-based workplace, but their utility at the individual level is limited due to large variability.

Using deep learning to generate synthetic B-mode musculoskeletal ultrasound images.

BACKGROUND AND OBJECTIVE: Deep learning approaches are common in image processing, but often rely on supervised learning, which requires a large volume of training images, usually accompanied by hand-crafted labels. As labelled data are often not available, it would be desirable to develop methods that allow such data to be compiled automatically. In this study, we used a Generative Adversarial Network (GAN) to generate realistic B-mode musculoskeletal ultrasound images, and tested the suitability of two automated labelling approaches. METHODS: We used a model including two GANs each trained to transfer an image from one domain to another. The two inputs were a set of 100 longitudinal images of the gastrocnemius medialis muscle, and a set of 100 synthetic segmented masks that featured two aponeuroses and a random number of 'fascicles'. The model output a set of synthetic ultrasound images and an automated segmentation of each real input image. This automated segmentation process was one of the two approaches we assessed. The second approach involved synthesising ultrasound images and then feeding these images into an ImageJ/Fiji-based automated algorithm, to determine whether it could detect the aponeuroses and muscle fascicles. RESULTS: Histogram distributions were similar between real and synthetic images, but synthetic images displayed less variation between samples and a narrower range. Mean entropy values were statistically similar (real: 6.97, synthetic: 7.03; p = 0.218), but the range was much narrower for synthetic images (6.91 - 7.11 versus 6.30 - 7.62). When comparing GAN-derived and manually labelled segmentations, intersection-over-union values- denoting the degree of overlap between aponeurosis labels- varied between 0.0280 - 0.612 (mean ± SD: 0.312 ± 0.159), and pennation angles were higher for the GAN-derived segmentations (25.1° vs. 19.3°; p < 0.001). For the second segmentation approach, the algorithm generally performed equally well on synthetic and real images, yielding pennation angles within the physiological range (13.8-20°). CONCLUSIONS: We used a GAN to generate realistic B-mode ultrasound images, and extracted muscle architectural parameters from these images automatically. This approach could enable generation of large labelled datasets for image segmentation tasks, and may also be useful for data sharing. Automatic generation and labelling of ultrasound images minimises user input and overcomes several limitations associated with manual analysis.

Role of Menopausal Transition and Physical Activity in Loss of Lean and Muscle Mass: A Follow-Up Study in Middle-Aged Finnish Women.

In midlife, women experience hormonal changes due to menopausal transition. A decrease especially in estradiol has been hypothesized to cause loss of muscle mass. This study investigated the effect of menopausal transition on changes in lean and muscle mass, from the total body to the muscle fiber level, among 47-55-year-old women. Data were used from the Estrogenic Regulation of Muscle Apoptosis (ERMA) study, where 234 women were followed from perimenopause to early postmenopause. Hormone levels (estradiol and follicle stimulating hormone), total and regional body composition (dual-energy X-ray absorptiometry (DXA) and computed tomography (CT) scans), physical activity level (self-reported and accelerometer-measured) and muscle fiber properties (muscle biopsy) were assessed at baseline and at early postmenopause. Significant decreases were seen in lean body mass (LBM), lean body mass index (LBMI), appendicular lean mass (ALM), appendicular lean mass index (ALMI), leg lean mass and thigh muscle cross-sectional area (CSA). Menopausal status was a significant predictor for all tested muscle mass variables, while physical activity was an additional significant contributor for LBM, ALM, ALMI, leg lean mass and relative muscle CSA. Menopausal transition was associated with loss of muscle mass at multiple anatomical levels, while physical activity was beneficial for the maintenance of skeletal muscle mass.

Muscle-tendon morphology and function following long-term exposure to repeated and strenuous mechanical loading.

We mapped structural and functional characteristics of muscle-tendon units in a population exposed to very long-term routine overloading. Twenty-eight military academy cadets (age = 21.00 ± 1.1 years; height = 176.1 ± 4.8 cm; mass = 73.8 ± 7.0 kg) exposed for over 24 months to repetitive overloading were profiled via ultrasonography with a senior subgroup of them (n = 11; age = 21.4 ± 1.0 years; height = 176.5 ± 4.8 cm; mass = 71.4 ± 6.6 kg) also tested while walking and marching on a treadmill. A group of eleven ethnicity- and age-matched civilians (age = 21.6 ± 0.7 years; height = 176.8 ± 4.3 cm; mass = 74.6 ± 5.6 kg) was also profiled and tested. Cadets and civilians exhibited similar morphology (muscle and tendon thickness and cross-sectional area, pennation angle, fascicle length) in 26 out of 29 sites including the Achilles tendon. However, patellar tendon thickness along the entire tendon was greater (P < .05) by a mean of 16% for the senior cadets compared with civilians. Dynamically, cadets showed significantly smaller ranges of fascicle length change and lower shortening velocity in medial gastrocnemius during walking (44.0% and 47.6%, P < .05-.01) and marching (27.5% and 34.3%, P < .05-.01) than civilians. Furthermore, cadets showed lower normalized soleus electrical activity during walking (22.7%, P < .05) and marching (27.0%, P < .05). Therefore, 24-36 months of continuous overloading, primarily occurring under aerobic conditions, leads to more efficient neural and mechanical behavior in the triceps surae complex, without any major macroscopic alterations in key anatomical structures.

Simple Muscle Architecture Analysis (SMA): An ImageJ macro tool to automate measurements in B-mode ultrasound scans.

In vivo measurements of muscle architecture (i.e. the spatial arrangement of muscle fascicles) are routinely included in research and clinical settings to monitor muscle structure, function and plasticity. However, in most cases such measurements are performed manually, and more reliable and time-efficient automated methods are either lacking completely, or are inaccessible to those without expertise in image analysis. In this work, we propose an ImageJ script to automate the entire analysis process of muscle architecture in ultrasound images: Simple Muscle Architecture Analysis (SMA). Images are filtered in the spatial and frequency domains with built-in commands and external plugins to highlight aponeuroses and fascicles. Fascicle dominant orientation is then computed in regions of interest using the OrientationJ plugin. Bland-Altman plots of analyses performed manually or with SMA indicate that the automated analysis does not induce any systematic bias and that both methods agree equally through the range of measurements. Our test results illustrate the suitability of SMA to analyse images from superficial muscles acquired with a broad range of ultrasound settings.

Effects of caffeine on neuromuscular function in a non-fatigued state and during fatiguing exercise.

NEW FINDINGS: What is the central question of the study? What are the effects of caffeine on neuromuscular function in a non-fatigued state and during fatiguing exercise? What is the main finding and its importance? In a non-fatigued state, caffeine decreased the duration of the silent period evoked by transcranial magnetic stimulation. Caffeine-induced reduction of inhibitory mechanisms in the central nervous system before exercise was associated with an increased performance. Individuals who benefit from caffeine ingestion may experience lower perception of effort during exercise and an accelerated recovery of M-wave amplitude postfatigue. This study elucidates the mechanisms of action of caffeine and demonstrates that inter-individual variability of its effects on neuromuscular function is a fruitful area for further work. ABSTRACT: Caffeine enhances exercise performance, but its mechanisms of action remain unclear. In this study, we investigated its effects on neuromuscular function in a non-fatigued state and during fatiguing exercise. Eighteen men participated in this randomized, double-blind, placebo-controlled crossover trial. Baseline measures included plantarflexion force, drop jump, squat jump, voluntary activation of triceps surae muscle, soleus muscle contractile properties, M-wave, α-motoneuron excitability (H-reflex), corticospinal excitability, short-interval intracortical inhibition, intracortical facilitation, silent period evoked by transcranial magnetic stimulation (SP) and plasma potassium and caffeine concentrations. Immediately after baseline testing, participants ingested caffeine (6 mg·kg-1 ) or placebo. After a 1-h rest, baseline measures were repeated, followed by a fatiguing stretch-shortening cycle exercise (sets of 40 bilateral rebound jumps on a sledge apparatus) until task failure. Neuromuscular testing was carried out throughout the fatigue protocol and afterwards. Caffeine enhanced drop jump height (by 4.2%) and decreased the SP (by 12.6%) in a non-fatigued state. A caffeine-related decrease in SP and short-interval intracortical inhibition before the fatiguing activity was associated with an increased time to task failure. The participants who benefitted from an improved performance on the caffeine day reported a significantly lower sense of effort during exercise and had an accelerated postexercise recovery of M-wave amplitude. Caffeine modulates inhibitory mechanisms of the CNS, recovery of M-wave amplitude and perception of effort. This study lays the groundwork for future examinations of differences in caffeine-induced neuromuscular changes between those who are deemed to benefit from caffeine ingestion and those who are not.

Individual Region- and Muscle-specific Hamstring Activity at Different Running Speeds.

INTRODUCTION: Hamstring strain injuries typically occur in the proximal biceps femoris long head (BFlh) at high running speeds. Strain magnitude seems to be the primary determinant of strain injury, and may be regulated by muscle activation. In running, BFlh strain is largest in the proximal region, especially at high speeds. However, region-specific activity has not been examined. This study examined the proximal-distal and intermuscular activity of BFlh and semitendinosus (ST) as a function of increasing running speed. METHODS: Thirteen participants ran at steady speeds of 4.1 (slow), 5.4 (moderate), and 6.8 m·s (fast) on a treadmill. Region- and muscle-specific EMG activity were recorded at each speed using high-density EMG, and were normalized to maximal voluntary isometric activity. Muscle-tendon unit lengths were calculated from kinematic recordings. Speed effects, regional, and intermuscular differences were tested with Statistical Parametric Mapping. RESULTS: With increasing running speed, EMG activity increased in all regions of both muscles to a similar extent in the clinically relevant late swing phase. Increases in muscle-tendon unit lengths in late swing as a function of running speed were comparatively small. In fast running, EMG activity was highest in late swing in all regions, and reached 115% ± 20% (proximal region, mean ± 95% confidence limit), 106% ± 11% (middle), and 124% ± 16% (distal) relative to maximal voluntary isometric activity in BFlh. Regional and intermuscular EMG patterns were highly individual, but each individual maintained similar proximal-distal and intermuscular EMG activity patterns across running speeds. CONCLUSIONS: Running is associated with highly individual hamstring activity patterns, but these patterns are similar across speeds. It may thus be crucial to implement running at submaximal speeds early after hamstring injury for restoration of normal neuromuscular function.