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., forS $$ S $$  = 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.

A deep learning approach for fast muscle water T2 mapping with subject specific fat T2 calibration from multi-spin-echo acquisitions

This work presents a deep learning approach for rapid and accurate muscle water T2 with subject-specific fat T2 calibration using multi-spin-echo acquisitions. This method addresses the computational limitations of conventional bi-component Extended Phase Graph fitting methods (nonlinear-least-squares and dictionary-based) by leveraging fully connected neural networks for fast processing with minimal computational resources. We validated the approach through in vivo experiments using two different MRI vendors. The results showed strong agreement of our deep learning approach with reference methods, summarized by Lin's concordance correlation coefficients ranging from 0.89 to 0.97. Further, the deep learning method achieved a significant computational time improvement, processing data 116 and 33 times faster than the nonlinear least squares and dictionary methods, respectively. In conclusion, the proposed approach demonstrated significant time and resource efficiency improvements over conventional methods while maintaining similar accuracy. This methodology makes the processing of water T2 data faster and easier for the user and will facilitate the utilization of the use of a quantitative water T2 map of muscle in clinical and research studies.

Compositional and Functional MRI of Skeletal Muscle: A Review

Due to its exceptional sensitivity to soft tissues, MRI has been extensively utilized to assess anatomical muscle parameters such as muscle volume and cross-sectional area. Quantitative Magnetic Resonance Imaging (qMRI) adds to the capabilities of MRI, by providing information on muscle composition such as fat content, water content, microstructure, hypertrophy, atrophy, as well as muscle architecture. In addition to compositional changes, qMRI can also be used to assess function for example by measuring muscle quality or through characterization of muscle deformation during passive lengthening/shortening and active contractions. The overall aim of this review is to provide an updated overview of qMRI techniques that can quantitatively evaluate muscle structure and composition, provide insights into the underlying biological basis of the qMRI signal, and illustrate how qMRI biomarkers of muscle health relate to function in healthy and diseased/injured muscles. While some applications still require systematic clinical validation, qMRI is now established as a comprehensive technique, that can be used to characterize a wide variety of structural and compositional changes in healthy and diseased skeletal muscle. Taken together, multiparametric muscle MRI holds great potential in the diagnosis and monitoring of muscle conditions in research and clinical applications. EVIDENCE LEVEL: 5 TECHNICAL EFFICACY: Stage 2.

Effects of fitting methods, high b-values and image quality on diffusion and perfusion quantification and reproducibility in the calf

PURPOSES

The study aimed to optimize diffusion-weighted imaging (DWI) image acquisition and analysis protocols in calf muscles by investigating the effects of different model-fitting methods, image quality, and use of high b-value and constraints on parameters of interest (POIs). The optimized modeling methods were used to select the optimal combinations of b-values, which will allow shorter acquisition time while achieving the same reliability as that obtained using 16 b-values.

METHODS

Test-retest baseline and high-quality DWI images of ten healthy volunteers were acquired on a 3T MR scanner, using 16 b-values, including a high b-value of 1200 s/mm², and structural T1-weighted images for calf muscle delineation. Three and six different fitting methods were used to derive ADC from monoexponential (ME) model and D_d, f_p, and D_p from intravoxel incoherent motion (IVIM) model, with or without the high b-value. The optimized ME and IVIM models were then used to determine the optimal combinations of b-values, obtainable with the least number of b-values, using the selection criteria of coefficient of variance (CV) ≤10% for all POIs.

RESULTS

The find minimum multivariate algorithm was more flexible and yielded smaller fitting errors. The 2-steps fitting method, with fixed D_d, performed the best for IVIM model. The inclusion of high b-value reduced outliers, while constraints improved 2-steps fitting only.

CONCLUSIONS

The optimal numbers of b-values for ME and IVIM models were nine and six b-values respectively. Test-retest reliability analyses showed that only ADC and D_d were reliable for calf diffusion evaluation, with CVs of 7.22% and 4.09%.

The effect of ageing on fat infiltration of thigh and paraspinal muscles in men

Background

Myosteatosis, skeletal muscle fat infiltration, is associated with inflammation and fibrosis. The age-related increase of myosteatosis is an important characteristic of sarcopenia and contributes to fragility.

Aims

To investigate the impact of healthy aging on intermuscular adipose tissue (IMAT) and muscle fat fraction (FF) in the thigh and the paraspinal muscles in males.

Methods

In 54 healthy males (age 20–70), all active hobby golfers, magnetic resonance imaging was performed to determine volume of IMAT, volume of muscle tissue (MT) and of percentage of FF.

Results

Between ages 20–70, at the thigh, IMAT/MT volume and MT FF increased annually by 2.9% and 1.3%, respectively. At the psoas IMAT/Psoas volume did not change with age. MT FF increased by 1.5% annually. At the erector spinae IMAT/Erector volume decreased by 0.3% and MT FF increased by 2.8% annually.

Discussion

With increasing age, in males, thigh muscle atrophied, muscle tissue was partly replaced by adipose tissue and remaining muscle tissue also contained more fat. Similar effects were observed in the erector spinae. The psoas muscle did not atrophy, although MT FF also increased with age. Overall correlations with age were weak to moderate with higher correlations observed in the paraspinal muscles.

Conclusions

Age-related increases of muscle fat infiltration were observed in the thigh and in the spine. Muscle atrophy did not occur in the psoas. In cross-sectional studies, an adjustment of volumetric parameters by muscle volume is advisable when comparing age-dependent results.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40520-022-02149-1.

Four functional magnetic resonance imaging techniques for skeletal muscle exploration, a systematic review

BACKGROUND

The study of muscle health has become more relevant lately, due to global aging and a higher incidence of musculoskeletal pathologies. Current exploration techniques, such as electromyography, do not provide accurate spatial information.

OBJECTIVE

The objective of this work is to perform a systematic review of the literature to synthesize the contributions that can offer functional MRI techniques commonly used in neuroimaging, applied to skeletal muscle: Blood Oxygen Level Dependent (BOLD), IntraVoxel Incoherent Motion (IVIM), Arterial Spin Labeling (ASL) and Dynamic Contrast Enhanced (DCE).

EVIDENCE ACQUISITION

Web of Science and Medline databases were searched, over the last 10 years, focused on the use of BOLD, ASL, IVIM or DCE in skeletal muscle.

EVIDENCE SYNTHESIS

59 articles were included after applying the selection criteria. 37 studies were performed in healthy subjects, and 22 in patients with different pathologies: in peripheral arterial disease, systemic sclerosis, diabetes, osteoporosis, adolescent idiopathic scoliosis, and dermatomyositis. Reference values in healthy subjects still vary in some cases.

CONCLUSION

The studies show the feasibility of implementing functional MRI through BOLD, ASL, IVIM or DCE imaging in several muscles and their possible utility in different pathologies. A synthesis of how to implement such exploration is given here.

CLINICAL IMPACT

These four techniques are based on sequences already present in clinical MRI scanners, therefore, their use for functional muscle exploration does not require additional investment. These techniques allow visualization and quantification of parameters associated with the vascular health of the muscles and represent interesting support for musculoskeletal exploration.

Conceptual Relationship Between Traditional Persian Medicine and Modern Nutrition in Obesity in Middle Age

Context: Over the last decades, the prevalence of overweight (BMI > 25) and obesity (BMI > 30) is being the most important health challenge in urban populations. The relationship between obesity and the individual’s temperament has not been studied yet. Evidence Acquisition: This review aimed to study the causes of obesity, especially in middle-aged people, according to the new evidence of conventional medicine and findings of Traditional Persian Medicine (TPM) physicians cited in their books. Databases including PubMed, Scopus, and Iran Medex were also searched with keywords obesity and overweight for recent evidence in conventional medicine. Results: Based on the traditional medicine findings, dystemprament or disequilibrium in Mizaj may promote obesity in middle-aged and elderly people. The attenuation of innate heat and intrinsic moisture in middle age could increase the chance of overweight and obesity. Recent studies in modern nutrition reveal a linear relationship between diminishing the basal metabolic rate and increasing age, especially in middle age. Accordingly, cold/wet-tempered people have more efficient energy homeostasis than hot/dry-tempered people that is similar to individuals with Firmicutes gut microbiota predominance. People living in high altitudes and cold-dry climates maintain their innate heat better than people living in wet climates, which is in accordance with the increasing brown adipose tissue thermogenesis in cold exposure, which decreases the chance of obesity. Conclusions: Based on traditional Persian medicine school, it is hypothesized that increasing age and diminishing innate heat besides the accumulation of phlegm (moisture) in the gastrointestinal tract of middle-aged individuals may be influential in altering gut microbiota and consequently obesity. It seems that there may be a correlation between cold/wet Mizaj and risk factors of obesity.

Intravoxel Incoherent Motion Magnetic Resonance Imaging in Skeletal Muscle: Review and Future Directions

Throughout the body, muscle structure and function can be interrogated using a variety of noninvasive magnetic resonance imaging (MRI) methods. Recently, intravoxel incoherent motion (IVIM) MRI has gained momentum as a method to evaluate components of blood flow and tissue diffusion simultaneously. Much of the prior research has focused on highly vascularized organs, including the brain, kidney, and liver. Unique aspects of skeletal muscle, including the relatively low perfusion at rest and large dynamic range of perfusion between resting and maximal hyperemic states, may influence the acquisition, postprocessing, and interpretation of IVIM data. Here, we introduce several of those unique features of skeletal muscle; review existing studies of IVIM in skeletal muscle at rest, in response to exercise, and in disease states; and consider possible confounds that should be addressed for muscle-specific evaluations. Most studies used segmented nonlinear least squares fitting with a b-value threshold of 200 sec/mm² to obtain IVIM parameters of perfusion fraction (f), pseudo-diffusion coefficient (D*), and diffusion coefficient (D). In healthy individuals, across all muscles, the average ± standard deviation of D was 1.46 ± 0.30 × 10^-3  mm² /sec, D* was 29.7 ± 38.1 × 10^-3  mm² /sec, and f was 11.1 ± 6.7%. Comparisons of reported IVIM parameters in muscles of the back, thigh, and leg of healthy individuals showed no significant difference between anatomic locations. Throughout the body, exercise elicited a positive change of all IVIM parameters. Future directions including advanced postprocessing models and potential sequence modifications are discussed. LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY: Stage 2.

Exercise-induced muscle damage: multi-parametric MRI quantitative assessment

BACKGROUND

To explore the value of magnetic resonance quantitative analysis using diffusion tensor imaging, T2 mapping, and intravoxel incoherent motion in the evaluation of eccentric exercise-induced muscle damage and to compare the effects of various eccentric exercise modes at different time points in rats.

METHODS

A total of 174 Sprague-Dawley male rats were randomly divided into five groups: control, once-only exercise, continuous exercise, intermittent exercise, and once-fatigue exercise groups. Each experimental group was divided into seven time-subgroups: 0.5 h, 24 h, 48 h, 72 h, 96 h, 120 h and 168 h after exercise. The quadriceps femoris muscles were then scanned using magnetic resonance imaging. The apparent diffusion coefficient and fractional anisotropy values of diffusion tensor imaging, T2 values of T2 mapping, D and D* values of intravoxel incoherent motion and optical density values of desmin were measured. Associations among different eccentric exercise programmes, magnetic resonance imaging findings, and histopathological results were evaluated. Dunnett's test, two-way repeated measures analysis of variance, and Pearson correlation analysis were used for statistical analysis.

RESULTS

Diffusion tensor imaging showed that the number of muscle fibre bundles decreased to varying degrees with different time points and eccentric exercises. Apparent diffusion coefficient values of the exercise groups showed a trend that first increased and then decreased, the opposite of fractional anisotropy. The specimens in all eccentric exercise programmes showed high signal T2 values after exercise, the highest among which was in the once-fatigue exercise group. D and D* in the experimental groups were significantly higher than those in the control group at 0.5-48 h after exercise. The apparent diffusion coefficient, fractional anisotropy, T2, D and D* values correlated with the optical density values of desmin.

CONCLUSIONS

Diffusion tensor imaging, T2 mapping, and intravoxel incoherent motion technology accurately reflect the degree of skeletal muscle damage and recovery associated with eccentric exercise. The degree of muscle damage was the lowest in the continuous exercise group and the highest in the once-fatigue exercise group, which may provide more information and guidance for the formulation of physical and athletic training programmes.

The effect of ageing on skeletal muscle as assessed by quantitative MR imaging: an association with frailty and muscle strength

Background

Skeletal muscles undergo changes with ageing which can cause sarcopenia that can result in frailty. Quantitative MRI may detect the muscle-deficit component of frailty which could help improve the understanding of ageing muscles.

Aims

To investigate whether quantitative MRI measures of T2, fat fraction (FF), diffusion tensor imaging and muscle volume can detect differences within the muscles between three age groups, and to assess how these measures compare with frailty index, gait speed and muscle power.

Methods

18 ‘young’ (18–30 years), 18 ‘middle-aged’ (31–68 years) and 18 ‘older’ (> 69 years) healthy participants were recruited. Participants had an MRI of their dominant thigh. Knee extension and flexion power and handgrip strength were measured. Frailty (English Longitudinal Study of Ageing frailty index) and gait speed were measured in the older participants.

Results

Young participants had a lower muscle MRI T2, FF and mean diffusivity than middle-aged and older participants; middle-aged participants had lower values than older participants. Young participants had greater muscle flexion and extension power, muscle volume and stronger hand grip than middle-aged and older participants; middle-aged participants had greater values than the older participants. Quantitative MRI measurements correlated with frailty index, gait speed, grip strength and muscle power.

Discussion

Quantitative MRI and strength measurements can detect muscle differences due to ageing. Older participants had raised T2, FF and mean diffusivity and lower muscle volume, grip strength and muscle power.

Conclusions

Quantitative MRI measurements correlate with frailty and muscle function and could be used for identifying differences across age groups within muscle.

Impact of Number of Segmented Tissues on SAR Prediction Accuracy in Deep Pelvic Hyperthermia Treatment Planning

Simple Summary

Hyperthermia treatment planning is the process of optimizing treatment quality using pre-treatment simulations. Although it has become a powerful tool, prediction accuracy is strongly dependent on the patient model. For deep hyperthermia in the pelvis, it is common that only four tissue categories are discriminated (bone, fat, muscle-like, and tumor). For the head and neck region, more tissues have been shown to be required for good prediction accuracy. Delineating is a labor-intensive and difficult process. Hence, it is important to find the optimum between accuracy and labor, but for deep pelvic hyperthermia, there are no published studies showing the impact of the number of tissues. We studied the trade-off between the segmentation detail needed and segmentation feasibility. Our findings indicate that including high water content tissues can impact simulation accuracy. Although our results, in general, underline the suitability of our current clinical protocol, they help to prioritize improvements for specific cases.

Abstract

In hyperthermia, the general opinion is that pre-treatment optimization of treatment settings requires a patient-specific model. For deep pelvic hyperthermia treatment planning (HTP), tissue models comprising four tissue categories are currently discriminated. For head and neck HTP, we found that more tissues are required for increasing accuracy. In this work, we evaluated the impact of the number of segmented tissues on the predicted specific absorption rate (SAR) for the pelvic region. Highly detailed anatomical models of five healthy volunteers were selected from a virtual database. For each model, seven lists with varying levels of segmentation detail were defined and used as an input for a modeling study. SAR changes were quantified using the change in target-to-hotspot-quotient and maximum SAR relative differences, with respect to the most detailed patient model. The main finding of this study was that the inclusion of high water content tissues in the segmentation may result in a clinically relevant impact on the SAR distribution and on the predicted hyperthermia treatment quality when considering our pre-established thresholds. In general, our results underline the current clinical segmentation protocol and help to prioritize any improvements.

Magnetic resonance elastography (MRE) shows significant reduction of thigh muscle stiffness in healthy older adults

Determining the effect of ageing on thigh muscle stiffness using magnetic resonance elastography (MRE) and investigate whether fat fraction and muscle cross-sectional area (CSA) are related to stiffness. Six healthy older adults in their eighth and ninth decade and eight healthy young men were recruited and underwent a 3 T MRI protocol including MRE and Dixon fat fraction imaging. Muscle stiffness, fat fraction and muscle CSA were calculated in ROIs corresponding to the four quadriceps muscles (i.e. vastus lateralis (VL), vastus medialis (VM), vastus intermedius (VI), rectus femoris (RF)), combined quadriceps, combined hamstrings and adductors and whole thigh. Muscle stiffness was significantly reduced (p < 0.05) in the older group in all measured ROIs except the VI (p = 0.573) and RF (p = 0.081). Similarly, mean fat fraction was significantly increased (p < 0.05) in the older group over all ROIs with the exception of the VI (p = 0.059) and VL muscle groups (p = 0.142). Muscle CSA was significantly reduced in older participants in the VM (p = 0.003) and the combined quadriceps (p = 0.001), hamstrings and adductors (p = 0.008) and whole thigh (p = 0.003). Over the whole thigh, stiffness was significantly negatively correlated with fat fraction (r = − 0.560, p = 0.037) and positively correlated with CSA (r = 0.749, p = 0.002). Stepwise regression analysis revealed that age was the most significant predictor of muscle stiffness (p = 0.001). These results suggest that muscle stiffness is significantly decreased in healthy older adults. Muscle fat fraction and muscle CSA are also significantly changed in older adults; however, age is the most significant predictor of muscle stiffness.

Normal values and test-retest variability of stimulated-echo diffusion tensor imaging and fat fraction measurements in the muscle

OBJECTIVES

To assess the test-retest variability of both diffusion parameters and fat fraction (FF) estimates in normal muscle, and to assess differences in normal values between muscles in the thigh.

METHODS

29 healthy volunteers (mean age 37 years, range 20-60 years, 17/29 males) completed the study. Magnetic resonance images of the mid-thigh were acquired using a stimulated echo acquisition mode-echoplanar imaging (STEAM-EPI) imaging sequence, to assess diffusion, and 2-point Dixon imaging, to assess FF. Imaging was repeated in 19 participants after a 30 min interval in order to assess test-retest variability of the measurements.

RESULTS

Intraclass correlation coefficients (ICCs) for test-retest variability were 0.99 [95% confidence interval, (CI): 0.98, 1] for FF, 0.94 (95% CI: 0.84, 0.97) for mean diffusivity and 0.89 (95% CI: 0.74, 0.96) for fractional anisotropy (FA). FF was higher in the hamstrings than the quadriceps by a mean difference of 1.81% (95% CI:1.63, 2.00)%, p < 0.001. Mean diffusivity was significantly lower in the hamstrings than the quadriceps (0.26 (0.13, 0.39) x10^-3 mm²s^-1, p < 0.001) whereas fractional anisotropy was significantly higher in the hamstrings relative to the quadriceps with a mean difference of 0.063 (0.05, 0.07), p < 0.001.

CONCLUSIONS

This study has shown excellent test-retest, variability in MR-based FF and diffusion measurements and demonstrated significant differences in these measures between hamstrings and quadriceps in the healthy thigh.

ADVANCES IN KNOWLEDGE

Test-retest variability is excellent for STEAM-EPI diffusion and 2-point Dixon-based FF measurements in the healthy muscle. Inter- and intraobserver variability were excellent for region of interest placement for STEAM-EPI diffusion and 2-point Dixon-based FF measurements in the healthy muscle. There are significant differences in FF and diffusion measurements between the hamstrings and quadriceps in the normal muscle.

Comparison Thigh Skeletal Muscles between Snowboarding Halfpipe Athletes and Healthy Volunteers Using Quantitative Multi-Parameter Magnetic Resonance Imaging at Rest

Background:

Magnetic resonance (MR) imaging provides a unique, noninvasive diagnostic platform to quantify the physiological and biochemical variables of skeletal muscle at rest. This study was to investigate the difference in thigh skeletal muscles between snowboarding halfpipe athletes and healthy volunteers via multiparametric MR imaging.

Methods:

A comparative study was conducted between 12 healthy volunteers and 14 snowboarding halfpipe athletes. MR scanning targeted the left leg at the level of the proximal thigh on a 3.0T MR system. The measured parameters compared between the two groups included T1, T2, T2* relaxation times, fat fraction (FF), and cross-sectional area (CSA) of the quadriceps femoris and the hamstring muscles. Statistical analysis was carried out using independent sample t-test. Interrater reliability was also assessed with intraclass correlation coefficients (ICCs).

Results:

It was statistically equivalent between two groups in age, body mass index, thigh circumference, calf circumference, systolic blood pressure, and resting heart rate (all P > 0.05). However, the T1 and T2 values of the hamstring muscles in the athlete group were found to be significantly shorter than those in control group (T1: 1063.3 ± 24.1 ms vs. 1112.0 ± 38.2 ms in biceps femoris, 1050.4 ± 31.2 ms vs. 1095.0 ± 39.5 ms in semitendinosus, 1053.1 ± 31.7 ms vs. 1118.4 ± 40.0 ms in semimembranosus, respectively; T2: 33.4 ± 0.7 ms vs. 36.1 ± 1.9 ms in biceps femoris, 34.6 ± 2.0 ms vs. 37.0 ± 1.9 ms in semitendinosus, 36.9 ± 1.5 ms vs. 38.9 ± 2.4 ms in semimembranosus, respectively; all P < 0.05) although T2* relaxation time was detected with no significant difference. The FF of the hamstring muscles was obviously less than the control group (5.5 ± 1.9% vs. 10.7 ± 4.7%, P < 0.001). In addition, the quadriceps' CSA in the athlete group was substantially larger than the control group (8039.0 ± 1072.3 vs. 6258.2 ± 852.0 mm², P < 0.001). Interrater reliability was excellent (ICC: 0.758–0.994).

Conclusion:

Multiple MR imaging parameters indicated significant differences between snowboarding halfpipe athletes and healthy volunteers in the thigh skeletal muscles.