MRI-based anatomical characterisation of lower-limb muscles in older women

Lower limb joint loading during high-impact activities: implication for bone health

Osteoporosis results in low-trauma fractures affecting millions globally, in particular elderly populations. Despite the inclusion of physical activity in fracture prevention strategies, the optimal bone-strengthening exercises remain uncertain, highlighting the need for a deeper understanding of lower limb joint loading dynamics across various exercise types and levels. This study examines lower limb joint loading during high-impact exercises across different intensities. A total of 40 healthy, active participants were recruited (mean ± SD: age of 40.3 ± 13.1 yr; height 1.71 ± 0.08 m; and mass 68.44 ± 11.67 kg). Motion capture data and ground reaction forces of 6 different exercises: a self-selected level of walking, running, countermovement jump, squat jump, unilateral hopping, and bilateral hopping were collected for each participant. Joint reaction forces were estimated using lower body musculoskeletal models developed in OpenSim. Running and hopping increased joint forces compared to walking, notably at the hip (83% and 21%), knee (134% and 94%), and ankle (94% and 77%), while jump exercises reduced hip and ankle loading compared to walking (36% and 19%). Joint loading varied with exercise type and intensity, with running faster increasing forces on all joints, particularly at the hip. Sprinting increased forces at the hip but lowered knee and ankle forces. Higher jumps intensified forces on all joints, while faster hopping reduced forces. The wide variation of lower limb joint loading observed across the exercises tested in this study underscores the importance of implementing diverse exercise routines to optimize overall bone health and strengthen the musculoskeletal structure. Practitioners must therefore ensure that exercise programs include movements that are specifically suitable for their intended purpose.

A novel mean shape based post-processing method for enhancing deep learning lower-limb muscle segmentation accuracy

This study aims at improving the lower-limb muscle segmentation accuracy of deep learning approaches based on Magnetic Resonance Imaging (MRI) scans, crucial for the diagnostic and therapeutic processes in musculoskeletal diseases. In general, segmentation methods such as U-Net deep learning neural networks can achieve good Dice Similarity Coefficient (DSC) values, e.g. around 0.83 to 0.91 on various cohorts. Some generic post-processing strategies have been studied to incorporate connectivity constraints into the resulting masks for the purpose of further improving the segmentation accuracy. In this paper, a novel mean shape (MS) based post-processing method is proposed, utilizing Statistical Shape Modelling (SSM) to fine-tune the segmentation output, taking into consideration the muscle anatomical shape. The methodology was compared to existing post-processing techniques and a commercial semi-automatic tool on MRI scans from two cohorts of post-menopausal women (10 Training, 8 Testing, voxel size 1.0x1.0x1.0 mm3). The MS based method obtained a mean DSC of 0.83 across the different analysed muscles and the best performance for the Hausdorff Distance (HD, 20.6 mm) and the Average Symmetric Surface Distance (ASSD, 2.1 mm). These findings highlight the feasibility and potential of using anatomical mean shape in post-processing of human lower-limb muscle segmentation task and indicate that the proposed method can be popularized to other biological organ segmentation mission.

Positional contrastive learning for improved thigh muscle segmentation in MR images

The accurate segmentation of individual muscles is essential for quantitative MRI analysis of thigh images. Deep learning methods have achieved state-of-the-art results in segmentation, but they require large numbers of labeled data to perform well. However, labeling individual thigh muscles slice by slice for numerous volumes is a laborious and time-consuming task, which limits the availability of annotated datasets. To address this challenge, self-supervised learning (SSL) emerges as a promising technique to enhance model performance by pretraining the model on unlabeled data. A recent approach, called positional contrastive learning, exploits the information given by the axial position of the slices to learn features transferable on the segmentation task. The aim of this work was to propose positional contrastive SSL for the segmentation of individual thigh muscles from MRI acquisitions in a population of elderly healthy subjects and to evaluate it on different levels of limited annotated data. An unlabeled dataset of 72 T1w MRI thigh acquisitions was available for SSL pretraining, while a labeled dataset of 52 volumes was employed for the final segmentation task, split into training and test sets. The effectiveness of SSL pretraining to fine-tune a U-Net architecture for thigh muscle segmentation was compared with that of a randomly initialized model (RND), considering an increasing number of annotated volumes (S = 1, 2, 5, 10, 20, 30, 40). Our results demonstrated that SSL yields substantial improvements in Dice similarity coefficient (DSC) when using a very limited number of labeled volumes (e.g., for S  = 1, DSC 0.631 versus 0.530 for SSL and RND, respectively). Moreover, enhancements are achievable even when utilizing the full number of labeled subjects, with DSC = 0.927 for SSL and 0.924 for RND. In conclusion, positional contrastive SSL was effective in obtaining more accurate thigh muscle segmentation, even with a very low number of labeled data, with a potential impact of speeding up the annotation process in clinics.

Automatic segmentation of lower limb muscles from MR images of post-menopausal women based on deep learning and data augmentation

Individual muscle segmentation is the process of partitioning medical images into regions representing each muscle. It can be used to isolate spatially structured quantitative muscle characteristics, such as volume, geometry, and the level of fat infiltration. These features are pivotal to measuring the state of muscle functional health and in tracking the response of the body to musculoskeletal and neuromusculoskeletal disorders. The gold standard approach to perform muscle segmentation requires manual processing of large numbers of images and is associated with significant operator repeatability issues and high time requirements. Deep learning-based techniques have been recently suggested to be capable of automating the process, which would catalyse research into the effects of musculoskeletal disorders on the muscular system. In this study, three convolutional neural networks were explored in their capacity to automatically segment twenty-three lower limb muscles from the hips, thigh, and calves from magnetic resonance images. The three neural networks (UNet, Attention UNet, and a novel Spatial Channel UNet) were trained independently with augmented images to segment 6 subjects and were able to segment the muscles with an average Relative Volume Error (RVE) between -8.6% and 2.9%, average Dice Similarity Coefficient (DSC) between 0.70 and 0.84, and average Hausdorff Distance (HD) between 12.2 and 46.5 mm, with performance dependent on both the subject and the network used. The trained convolutional neural networks designed, and data used in this study are openly available for use, either through re-training for other medical images, or application to automatically segment new T1-weighted lower limb magnetic resonance images captured with similar acquisition parameters.

Automatic segmentation of skeletal muscles from MR images using modified U-Net and a novel data augmentation approach

Rapid and accurate muscle segmentation is essential for the diagnosis and monitoring of many musculoskeletal diseases. As gold standard, manual annotation suffers from intensive labor and high inter-operator reproducibility errors. In this study, deep learning (DL) based automatic muscle segmentation from MR scans is investigated for post-menopausal women, who normally experience a decline in muscle volume. The performance of four Deep Learning (DL) models was evaluated: U-Net and UNet++ and two modified U-Net networks, which combined feature fusion and attention mechanisms (Feature-Fusion-UNet, FFU, and Attention-Feature-Fusion-UNet, AFFU). The models were tested for automatic segmentation of 16-lower limb muscles from MRI scans of two cohorts of post-menopausal women (11 subjects in PMW-1, 8 subjects in PMW-2; from two different studies so considered independent datasets) and 10 obese post-menopausal women (PMW-OB). Furthermore, a novel data augmentation approach is proposed to enlarge the training dataset. The results were assessed and compared by using the Dice similarity coefficient (DSC), relative volume error (RVE), and Hausdorff distance (HD). The best performance among all four DL models was achieved by AFFU (PMW-1: DSC 0.828 ± 0.079, 1-RVE 0.859 ± 0.122, HD 29.9 mm ± 26.5 mm; PMW-2: DSC 0.833 ± 0.065, 1-RVE 0.873 ± 0.105, HD 25.9 mm ± 27.9 mm; PMW-OB: DSC 0.862 ± 0.048, 1-RVE 0.919 ± 0.076, HD 34.8 mm ± 46.8 mm). Furthermore, the augmentation of data significantly improved the DSC scores of U-Net and AFFU for all 16 tested muscles (between 0.23% and 2.17% (DSC), 1.6%-1.93% (1-RVE), and 9.6%-19.8% (HD) improvement). These findings highlight the feasibility of utilizing DL models for automatic segmentation of muscles in post-menopausal women and indicate that the proposed augmentation method can enhance the performance of models trained on small datasets.

Effect of Muscle Forces on Femur During Level Walking Using a Virtual Population of Older Women

Aging is associated with a greater risk of muscle and bone disorders such as sarcopenia and osteoporosis. These conditions substantially affect one's mobility and quality of life. In the past, muscles and bones are often studied separately using generic or scaled information that are not personal-specific, nor are they representative of the large variations seen in the elderly population. Consequently, the mechanical interaction between the aged muscle and bone is not well understood, especially when carrying out daily activities. This study presents a coupling approach across the body and the organ level, using fully personal-specific musculoskeletal and finite element models in order to study femoral loading during level walking. Variations in lower limb muscle volume/force were examined using a virtual population. These muscle forces were then applied to the finite element model of the femur to study the variations in predicted strains. The study shows that effective coupling across two scales can be carried out to study the muscle-bone interaction in elderly women. The generation of a virtual population is a feasible approach to augment anatomical variations based on a small population that could mimic variations seen in a larger cohort. This is a valuable alternative to overcome the limitation or the need to collect dataset from a large population, which is both time and resource consuming.

Comparative validation of two patient-specific modelling pipelines for predicting knee joint forces during level walking

Over the past few years, the use of computer models and simulations tailored to the patient's physiology to assist clinical decision-making has increased enormously.While several pipelines to develop personalized models exist, their adoption on a large scale is still limited due to the required niche computational skillset and the lengthy operations required. Novel toolboxes, such as STAPLE, promise to streamline and expedite the development of image-based skeletal lower limb models. STAPLE-generated models can be rapidly generated, with minimal user input, and present similar joint kinematics and kinetics compared to models developed employing the established INSIGNEO pipeline. Yet, it is unclear how much the observed discrepancies scale up and affect joint contact force predictions. In this study, we compared image-based musculoskeletal models developed (i) with the INSIGNEO pipeline and (ii) with a semi-automated pipeline that combines STAPLE and nmsBuilder, and assessed their accuracy against experimental implant data.Our results showed that both pipelines predicted similar total knee joint contact forces between one another in terms of profiles and average values, characterized by a moderately high level of agreement with the experimental data. Nonetheless, the Student t-test revealed statistically significant differences between both pipelines. Of note, the STAPLE-based pipeline required considerably less time than the INSIGNEO pipeline to generate a musculoskeletal model (i.e., 60 vs 160 min). This is likely to open up opportunities for the use of personalized musculoskeletal models in clinical practice, where time is of the essence.

Deformable image registration based on single or multi-atlas methods for automatic muscle segmentation and the generation of augmented imaging datasets

Muscle segmentation is a process relied upon to gather medical image-based muscle characterisation, useful in directly assessing muscle volume and geometry, that can be used as inputs to musculoskeletal modelling pipelines. Manual or semi-automatic techniques are typically employed to segment the muscles and quantify their properties, but they require significant manual labour and incur operator repeatability issues. In this study an automatic process is presented, aiming to segment all lower limb muscles from Magnetic Resonance (MR) imaging data simultaneously using three-dimensional (3D) deformable image registration (single inputs or multi-atlas). Twenty-three of the major lower limb skeletal muscles were segmented from five subjects, with an average Dice similarity coefficient of 0.72, and average absolute relative volume error (RVE) of 12.7% (average relative volume error of -2.2%) considering the optimal subject combinations. The multi-atlas approach showed slightly better accuracy (average DSC: 0.73; average RVE: 1.67%). Segmented MR imaging datasets of the lower limb are not widely available in the literature, limiting the potential of new, probabilistic methods such as deep learning to be used in the context of muscle segmentation. In this work, Non-linear deformable image registration is used to generate 69 manually checked, segmented, 3D, artificial datasets, allowing access for future studies to use these new methods, with a large amount of reliable reference data.

Ultrasound Measurement of Rectus Femoris Muscle Parameters for Discriminating Sarcopenia in Community-Dwelling Adults

OBJECTIVES

Sarcopenia patients require more medical attention and caretaking. As such, early detection of sarcopenia and appropriate interventions are crucial for decreasing medical costs and meeting the challenges of aging populations. The aim of the present study was to develop a reliable and accurate model to estimate muscle mass using ultrasound-derived parameters from the rectus femoris (RF), referenced by dual-energy X-ray absorptiometry.

METHODS

Cross-sectional study was performed. The study patients were recruited by Taipei Veterans General Hospital (No. 2016-07-013C) between 2016 and 2019. A total of 91 community-dwelling adults (35 men and 56 women) were enrolled in this study. Ultrasound measurements of RF muscle thickness (MT), cross-sectional area (CSA), and muscle volume (MV) were performed in B-mode. Muscle strength and physical performance were also examined. Multivariate linear regression was used to build models for the prediction of appendicular skeletal muscle index (ASMI) based on MT, CSA, and MV values. The accuracy of ultrasound RF measurements for predicting sarcopenia was evaluated by using receiver operating characteristic (ROC) curve analysis.

RESULTS

The regression equations used for ASMI prediction (adjusted body mass index, sex, and leg length) had high precision and low error. Moreover, the MV model results were close to those of the CSA model and higher than those of the MT model. The ROC analysis showed that both MV and CSA had excellent discrimination when assessing sarcopenia (AUC = 0.83 and 0.81, respectively), whereas MT showed acceptable discrimination (AUC = 0.73).

CONCLUSIONS

Ultrasound-derived RF MV was accurate when predicting ASMI and diagnosing sarcopenia in community-dwelling adults.

Towards defining muscular regions of interest from axial magnetic resonance imaging with anatomical cross-reference: a scoping review of lateral hip musculature

Background

Measures of hip muscle morphology and composition (e.g., muscle size and fatty infiltration) are possible with magnetic resonance imaging (MRI). Standardised protocols or guidelines do not exist for evaluation of hip muscle characteristics, hindering reliable and valid inter-study analysis. This scoping review aimed to collate and synthesise MRI methods for measuring lateral hip muscle size and fatty infiltration to inform the future development of standardised protocols.

Methods

Five electronic databases (Medline, CINAHL, Embase, SportsDISCUS and AMED) were searched. Healthy or musculoskeletal pain populations that used MRI to assess lateral hip muscle size and fatty infiltration were included. Lateral hip muscles of interest included tensor fascia late (TFL), gluteus maximus, gluteus medius, and gluteus minimus. Data on MRI parameters, axial slice location, muscle size and fatty infiltrate measures were collected and analysed. Cross referencing for anatomical locations were made between MRI axial slice and E-12 anatomical plastinate sections.

Results

From 2684 identified publications, 78 studies contributed data on volume (n = 31), cross sectional area (CSA) (n = 24), and fatty infiltration (n = 40). Heterogeneity was observed for MRI parameters and anatomical boundaries scrutinizing hip muscle size and fatty infiltration. Seven single level axial slices were identified that provided consistent CSA measurement, including three for both gluteus maximus and TFL, and four for both gluteus medius and minimus. For assessment of fatty infiltration, six axial slice locations were identified including two for TFL, and four for each of the gluteal muscles.

Conclusions

Several consistent anatomical levels were identified for single axial MR slice to facilitate muscle size and fatty infiltration muscle measures at the hip, providing the basis for reliable and accurate data synthesis and improvements in the validity of future between studies analyses. This work establishes the platform for standardised methods for the MRI assessment of lateral hip musculature and will aid in the examination of musculoskeletal conditions around the hip joint. Further studies into whole muscle measures are required to further optimise methodological parameters for hip muscle assessment.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12891-022-05439-x.

Uncertainty in Muscle–Tendon Parameters can Greatly Influence the Accuracy of Knee Contact Force Estimates of Musculoskeletal Models

Understanding the sources of error is critical before models of the musculoskeletal system can be usefully translated. Using in vivo measured tibiofemoral forces, the impact of uncertainty in muscle–tendon parameters on the accuracy of knee contact force estimates of a generic musculoskeletal model was investigated following a probabilistic approach. Population variability was introduced to the routine musculoskeletal modeling framework by perturbing input parameters of the lower limb muscles around their baseline values. Using ground reaction force and skin marker trajectory data collected from six subjects performing body-weight squat, the knee contact force was calculated for the perturbed models. The combined impact of input uncertainties resulted in a considerable variation in the knee contact force estimates (up to 2.1 BW change in the predicted force), especially at larger knee flexion angles, hence explaining up to 70% of the simulation error. Although individual muscle groups exhibited different contributions to the overall error, variation in the maximum isometric force and pathway of the muscles showed the highest impacts on the model outcomes. Importantly, this study highlights parameters that should be personalized in order to achieve the best possible predictions when using generic musculoskeletal models for activities involving deep knee flexion.

Paggiosi M, McCloskey EV, Viceconti M, Mazzà C, Li X. Femoral neck strain prediction during level walking using a combined musculoskeletal and finite element model approach

Recently, coupled musculoskeletal-finite element modelling approaches have emerged as a way to investigate femoral neck loading during various daily activities. Combining personalised gait data with finite element models will not only allow us to study changes in motion/movement, but also their effects on critical internal structures, such as the femur. However, previous studies have been hampered by the small sample size and the lack of fully personalised data in order to construct the coupled model. Therefore, the aim of this study was to build a pipeline for a fully personalised multiscale (body-organ level) model to investigate the strain levels at the femoral neck during a normal gait cycle. Five postmenopausal women were included in this study. The CT and MRI scans of the lower limb, and gait data were collected for all participants. Muscle forces derived from the body level musculoskeletal models were used as boundary constraints on the finite element femur models. Principal strains were estimated at the femoral neck region during a full gait cycle. Considerable variation was found in the predicted peak strain among individuals with mean peak first principal strain of 0.24% ± 0.11% and mean third principal strain of -0.29% ± 0.24%. For four individuals, two overall peaks of the maximum strains were found to occur when both feet were in contact with the floor, while one individual had one peak at the toe-off phase. Both the joint contact forces and the muscular forces were found to substantially influence the loading at the femoral neck. A higher correlation was found between the predicted peak strains and the gluteus medius (R² ranged between 0.95 and 0.99) than the hip joint contact forces (R² ranged between 0.63 and 0.96). Therefore, the current findings suggest that personal variations are substantial, and hence it is important to consider multiple subjects before deriving general conclusions for a target population.