Self-supervised learning for human activity recognition using 700,000 person-days of wearable data.

Accurate physical activity monitoring is essential to understand the impact of physical activity on one's physical health and overall well-being. However, advances in human activity recognition algorithms have been constrained by the limited availability of large labelled datasets. This study aims to leverage recent advances in self-supervised learning to exploit the large-scale UK Biobank accelerometer dataset-a 700,000 person-days unlabelled dataset-in order to build models with vastly improved generalisability and accuracy. Our resulting models consistently outperform strong baselines across eight benchmark datasets, with an F1 relative improvement of 2.5-130.9% (median 24.4%). More importantly, in contrast to previous reports, our results generalise across external datasets, cohorts, living environments, and sensor devices. Our open-sourced pre-trained models will be valuable in domains with limited labelled data or where good sampling coverage (across devices, populations, and activities) is hard to achieve.

Device-measured movement behaviours in over 20,000 China Kadoorie Biobank participants.

BACKGROUND: Movement behaviours, including physical activity, sedentary behaviour, and sleep have been shown to be associated with several chronic diseases. However, they have not been objectively measured in large-scale prospective cohort studies in low-and middle-income countries. We aim to describe the patterns of device-measured movement behaviours collected in the China Kadoorie Biobank (CKB) study. METHODS: During 2020 and 2021, a random subset of 25,087 surviving CKB individuals participated in the 3rd resurvey of the CKB. Among them, 22,511 (89.7%) agreed to wear an Axivity AX3 wrist-worn triaxial accelerometer for seven consecutive days to assess their habitual movement behaviours. We developed a machine-learning model to infer time spent in four movement behaviours [i.e. sleep, sedentary behaviour, light intensity physical activity (LIPA), and moderate-to-vigorous physical activity (MVPA)]. Descriptive analyses were performed for wear-time compliance and patterns of movement behaviours by different participant characteristics. RESULTS: Data from 21,897 participants (aged 65.4 ± 9.1 years; 35.4% men) were received for demographic and wear-time analysis, with a median wear-time of 6.9 days (IQR: 6.1-7.0). Among them, 20,370 eligible participants were included in movement behavior analyses. On average, they had 31.1 mg/day (total acceleration) overall activity level, accumulated 7.7 h/day (32.3%) of sleep time, 8.8 h/day (36.6%) sedentary, 5.7 h/day (23.9%) in light physical activity, and 104.4 min/day (7.2%) in moderate-to-vigorous physical activity. There was an inverse relationship between age and overall acceleration with an observed decline of 5.4 mg/day (17.4%) per additional decade. Women showed a higher activity level than men (32.3 vs 28.8 mg/day) and there was a marked geographical disparity in the overall activity level and time allocation. CONCLUSIONS: This is the first large-scale accelerometer data collected among Chinese adults, which provides rich and comprehensive information about device-measured movement behaviour patterns. This resource will enhance our knowledge about the potential relevance of different movement behaviours for chronic disease in Chinese adults.

Genome-Wide DNA Methylation Profiling as Frontline Diagnostics for Central Nervous System Embryonal Tumors in Hong Kong.

This paper examines the link between CNS tumor biology and heterogeneity and the use of genome-wide DNA methylation profiling as a clinical diagnostic platform. CNS tumors are the most common solid tumors in children, and their prognosis remains poor. This study retrospectively analyzed pediatric patients with CNS embryonal tumors in Hong Kong between 1999 and 2017, using data from the territory-wide registry and available formalin-fixed paraffin-embedded tumor tissue. After processing archival tumor tissue via DNA extraction, quantification, and methylation profiling, the data were analyzed by using the web-based DKFZ classifier (Molecular Neuropathology (MNP) 2.0 v11b4) and t-SNE analysis. Methylation profiles were deemed informative in 85 samples. Epigenetic data allowed molecular subgrouping and confirmed diagnosis in 65 samples, verified histologic diagnosis in 8, and suggested an alternative diagnosis in 12. This study demonstrates the potential of DNA methylation profiling in characterizing pediatric CNS embryonal tumors in a large cohort from Hong Kong, which should enable regional and international collaboration in future pediatric neuro-oncology research.

Adipose tissue-derived human mesenchymal stromal cells can better suppress complement lysis, engraft and inhibit acute graft-versus-host disease in mice.

BACKGROUND: Acute graft-versus-host disease (aGvHD) is a life-threatening complication of allogeneic hematopoietic stem cell transplantation (HSCT). Transplantation of immunosuppressive human mesenchymal stromal cells (hMSCs) can protect against aGvHD post-HSCT; however, their efficacy is limited by poor engraftment and survival. Moreover, infused MSCs can be damaged by activated complement, yet strategies to minimise complement injury of hMSCs and improve their survival are limited. METHODS: Human MSCs were derived from bone marrow (BM), adipose tissue (AT) and umbilical cord (UC). In vitro immunomodulatory potential was determined by co-culture experiments between hMSCs and immune cells implicated in aGvHD disease progression. BM-, AT- and UC-hMSCs were tested for their abilities to protect aGvHD in a mouse model of this disease. Survival and clinical symptoms were monitored, and target tissues of aGvHD were examined by histopathology and qPCR. Transplanted cell survival was evaluated by cell tracing and by qPCR. The transcriptome of BM-, AT- and UC-hMSCs was profiled by RNA-sequencing. Focused experiments were performed to compare the expression of complement inhibitors and the abilities of hMSCs to resist complement lysis. RESULTS: Human MSCs derived from three tissues divergently protected against aGvHD in vivo. AT-hMSCs preferentially suppressed complement in vitro and in vivo, resisted complement lysis and survived better after transplantation when compared to BM- and UC-hMSCs. AT-hMSCs also prolonged survival and improved the symptoms and pathological features of aGvHD. We found that complement-decay accelerating factor (CD55), an inhibitor of complement, is elevated in AT-hMSCs and contributed to reduced complement activation. We further report that atorvastatin and erlotinib could upregulate CD55 and suppress complement in all three types of hMSCs. CONCLUSION: CD55, by suppressing complement, contributes to the improved protection of AT-hMSCs against aGvHD. The use of AT-hMSCs or the upregulation of CD55 by small molecules thus represents promising new strategies to promote hMSC survival to improve the efficacy of transplantation therapy. As complement injury is a barrier to all types of hMSC therapy, our findings are of broad significance to enhance the use of hMSCs for the treatment of a wide range of disorders.

Development and Validation of a Machine Learning Wrist-worn Step Detection Algorithm with Deployment in the UK Biobank.

Background: Step count is an intuitive measure of physical activity frequently quantified in a range of health-related studies; however, accurate quantification of step count can be difficult in the free-living environment, with step counting error routinely above 20% in both consumer and research-grade wrist-worn devices. This study aims to describe the development and validation of step count derived from a wrist-worn accelerometer and to assess its association with cardiovascular and all-cause mortality in a large prospective cohort study. Methods: We developed and externally validated a hybrid step detection model that involves self-supervised machine learning, trained on a new ground truth annotated, free-living step count dataset (OxWalk, n=39, aged 19-81) and tested against other open-source step counting algorithms. This model was applied to ascertain daily step counts from raw wrist-worn accelerometer data of 75,493 UK Biobank participants without a prior history of cardiovascular disease (CVD) or cancer. Cox regression was used to obtain hazard ratios and 95% confidence intervals for the association of daily step count with fatal CVD and all-cause mortality after adjustment for potential confounders. Findings: The novel step algorithm demonstrated a mean absolute percent error of 12.5% in free-living validation, detecting 98.7% of true steps and substantially outperforming other recent wrist-worn, open-source algorithms. Our data are indicative of an inverse dose-response association, where, for example, taking 6,596 to 8,474 steps per day was associated with a 39% [24-52%] and 27% [16-36%] lower risk of fatal CVD and all-cause mortality, respectively, compared to those taking fewer steps each day. Interpretation: An accurate measure of step count was ascertained using a machine learning pipeline that demonstrates state-of-the-art accuracy in internal and external validation. The expected associations with CVD and all-cause mortality indicate excellent face validity. This algorithm can be used widely for other studies that have utilised wrist-worn accelerometers and an open-source pipeline is provided to facilitate implementation.

Chronic heart failure following hemorrhagic myocardial infarction: mechanism, treatment and outlook.

Myocardial infarction (MI), the blockage of arterial blood supply of the heart, is among the most common causes of death worldwide. Even when patients receive immediate treatment by re-opening blocked arteries, they often develop chronic heart failure (CHF) in the aftermath of MI events. Yet, the factors that contribute to the development of MI-associated CHF are poorly understood. In our recent study (Nat Commun 13:6394), we link intramyocardial hemorrhage, an injury which can occur during reperfusion of areas affected by MI, to an increased risk of CHF. Mechanistically, our data suggest that an iron-induced adverse cascade of events after hemorrhagic MI drives fatty degeneration of infarcted tissue, which ultimately contributes to negative cardiac remodeling. In this Microreview, we discuss the implications of our findings regarding the molecular mechanism, more targeted treatment options as well as perspectives in the clinical care of CHF after hemorrhagic MI.

Amount and intensity of physical activity and risk of incident cancer in the UK Biobank.

Importance: The influence of total daily and light intensity activity on cancer risk remains unclear, as most existing knowledge is drawn from studies relying on self-reported leisure-time activities of moderate-vigorous intensity. Objective: To investigate associations between total daily activity, including step counts, and activity intensity on incident cancer risk. Design Setting and Participants: Prospective analysis of cancer-free UK Biobank participants who wore accelerometers for 7-days (between 2013-2015), followed for cancer incidence through national registries (mean follow-up 5.8 years (SD=1.3)). Exposures: Time-series machine learning models derived daily total activity (average acceleration), behaviour time, step counts, and peak 30-minute cadence from wrist-based accelerometer data. Main Outcomes and Measures: A composite cancer outcome of 13 cancers previously associated with low physical activity (bladder, breast, colon, endometrial, oesophageal adenocarcinoma, gastric cardia, head and neck, kidney, liver, lung, myeloid leukaemia, myeloma, and rectum) based on previous studies of self-reported activity. Cox proportional hazards regression models estimated hazard ratios (HR) and 95% confidence intervals (CI), adjusted for age, sex, ethnicity, smoking, alcohol, education, Townsend Deprivation Index, and reproductive factors. Associations of reducing sedentary time in favour of increased light and moderate-vigorous activity were examined using compositional data analyses. Results: Among 86 556 participants (mean age 62.0 years (SD=7.9) at accelerometer assessment), 2 669 cancers occurred. Higher total physical activity was associated with a lower overall cancer risk (HR1SD=0.85, [95%CI 0.81-0.89]). On average, reallocating one hour/day from sedentary behaviour to moderate-vigorous physical activity was associated with a lower risk (HR=0.92, [0.89-0.95]), as was reallocating one hour/day to light-intensity physical activity (HR=0.94, [0.92-0.96]). Compared to individuals taking 5 000 daily steps, those who took 9 000 steps had an 18% lower risk of physical-activity-related cancer (HR=0.82, [0.74-0.90]). We found no significant association with peak 30-minute cadence after adjusting for total steps. Conclusion and Relevance: Higher total daily physical activity and less sedentary time, in favour of both light and moderate-vigorous intensity activity, were associated with a lower risk of certain cancers. For less active adults, increasing step counts by 4 000 daily steps may be a practical public health intervention for lowering the risk of some cancers.

Intramyocardial hemorrhage drives fatty degeneration of infarcted myocardium.

Sudden blockage of arteries supplying the heart muscle contributes to millions of heart attacks (myocardial infarction, MI) around the world. Although re-opening these arteries (reperfusion) saves MI patients from immediate death, approximately 50% of these patients go on to develop chronic heart failure (CHF) and die within a 5-year period; however, why some patients accelerate towards CHF while others do not remains unclear. Here we show, using large animal models of reperfused MI, that intramyocardial hemorrhage - the most damaging form of reperfusion injury (evident in nearly 40% of reperfused ST-elevation MI patients) - drives delayed infarct healing and is centrally responsible for continuous fatty degeneration of the infarcted myocardium contributing to adverse remodeling of the heart. Specifically, we show that the fatty degeneration of the hemorrhagic MI zone stems from iron-induced macrophage activation, lipid peroxidation, foam cell formation, ceroid production, foam cell apoptosis and iron recycling. We also demonstrate that timely reduction of iron within the hemorrhagic MI zone reduces fatty infiltration and directs the heart towards favorable remodeling. Collectively, our findings elucidate why some, but not all, MIs are destined to CHF and help define a potential therapeutic strategy to mitigate post-MI CHF independent of MI size.

Clinical translation of patient-derived tumour organoids- bottlenecks and strategies.

Multiple three-dimensional (3D) tumour organoid models assisted by multi-omics and Artificial Intelligence (AI) have contributed greatly to preclinical drug development and precision medicine. The intrinsic ability to maintain genetic and phenotypic heterogeneity of tumours allows for the reconciliation of shortcomings in traditional cancer models. While their utility in preclinical studies have been well established, little progress has been made in translational research and clinical trials. In this review, we identify the major bottlenecks preventing patient-derived tumour organoids (PDTOs) from being used in clinical setting. Unsuitable methods of tissue acquisition, disparities in establishment rates and a lengthy timeline are the limiting factors for use of PDTOs in clinical application. Potential strategies to overcome this include liquid biopsies via circulating tumour cells (CTCs), an automated organoid platform and optical metabolic imaging (OMI). These proposed solutions accelerate and optimize the workflow of a clinical organoid drug screening. As such, PDTOs have the potential for potential applications in clinical oncology to improve patient outcomes. If remarkable progress is made, cancer patients can finally benefit from this revolutionary technology.

Reallocation of time between device-measured movement behaviours and risk of incident cardiovascular disease.

OBJECTIVE: To improve classification of movement behaviours in free-living accelerometer data using machine-learning methods, and to investigate the association between machine-learned movement behaviours and risk of incident cardiovascular disease (CVD) in adults. METHODS: Using free-living data from 152 participants, we developed a machine-learning model to classify movement behaviours (moderate-to-vigorous physical activity behaviours (MVPA), light physical activity behaviours, sedentary behaviour, sleep) in wrist-worn accelerometer data. Participants in UK Biobank, a prospective cohort, were asked to wear an accelerometer for 7 days, and we applied our machine-learning model to classify their movement behaviours. Using compositional data analysis Cox regression, we investigated how reallocating time between movement behaviours was associated with CVD incidence. RESULTS: In leave-one-participant-out analysis, our machine-learning method classified free-living movement behaviours with mean accuracy 88% (95% CI 87% to 89%) and Cohen's kappa 0.80 (95% CI 0.79 to 0.82). Among 87 498 UK Biobank participants, there were 4105 incident CVD events. Reallocating time from any behaviour to MVPA, or reallocating time from sedentary behaviour to any behaviour, was associated with lower CVD risk. For an average individual, reallocating 20 min/day to MVPA from all other behaviours proportionally was associated with 9% (95% CI 7% to 10%) lower risk, while reallocating 1 hour/day to sedentary behaviour from all other behaviours proportionally was associated with 5% (95% CI 3% to 7%) higher risk. CONCLUSION: Machine-learning methods classified movement behaviours accurately in free-living accelerometer data. Reallocating time from other behaviours to MVPA, and from sedentary behaviour to other behaviours, was associated with lower risk of incident CVD, and should be promoted by interventions and guidelines.

Early Development of Colonic Adenocarcinoma With Minimal Polyposis in a Young Child With Metastatic Hepatoblastoma and Germline APC Mutation.

Germline adenomatous polyposis coli (APC) gene mutation is a cancer-predisposing condition commonly presenting as familial adenomatous polyposis. We describe a patient first diagnosed at the age of 3 years with metastatic hepatoblastoma. With a positive family history, germline testing confirmed maternally inherited APC mutation (p.Thr899Ansfs*13). The patient was subsequently diagnosed at 8 years with colonic adenocarcinoma in the absence of macroscopic polyposis. Total colectomy with adjuvant chemotherapy was delivered and the patient remained disease-free for 5 years since the second diagnosis. This report demonstrates the importance of considering germline APC mutation in children with hepatoblastoma, who may benefit from the early institution of colonoscopic surveillance.

Mesenchymal stem cells accelerated growth and metastasis of neuroblastoma and preferentially homed towards both primary and metastatic loci in orthotopic neuroblastoma model.

BACKGROUND: Majority of neuroblastoma patients develop metastatic disease at diagnosis and their prognosis is poor with current therapeutic approach. Major challenges are how to tackle the mechanisms responsible for tumorigenesis and metastasis. Human mesenchymal stem cells (hMSCs) may be actively involved in the constitution of cancer microenvironment. METHODS: An orthotopic neuroblastoma murine model was utilized to mimic the clinical scenario. Human neuroblastoma cell line SK-N-LP was transfected with luciferase gene, which were inoculated with/without hMSCs into the adrenal area of SCID-beige mice. The growth and metastasis of neuroblastoma was observed by using Xenogen IVIS 100 in vivo imaging and evaluating gross tumors ex vivo. The homing of hMSCs towards tumor was analyzed by tracing fluorescence signal tagged on hMSCs using CRI Maestro™ imaging system. RESULTS: hMSCs mixed with neuroblastoma cells significantly accelerated tumor growth and apparently enhanced metastasis of neuroblastoma in vivo. hMSCs could be recruited by primary tumor and also become part of the tumor microenvironment in the metastatic lesion. The metastatic potential was consistently reduced in lung and tumor when hMSCs were pre-treated with stromal cell derived factor-1 (SDF-1) blocker, AMD3100, suggesting that the SDF-1/CXCR4 axis was one of the prime movers in the metastatic process. CONCLUSIONS: hMSCs accelerated and facilitated tumor formation, growth and metastasis. Furthermore, the homing propensity of hMSCs towards both primary tumor and metastatic loci can also provide new therapeutic insights in utilizing bio-engineered hMSCs as vehicles for targeted anti-cancer therapy.

Bismuth Porphyrin Antagonizes Cisplatin-Induced Nephrotoxicity via Unexpected Metallothionein-Independent Mechanisms.

Cisplatin (CDDP) has been a highly successful anticancer drug in cancer therapy; however, its further application suffers severe nephrotoxicity. Herein, we identify bismuth tetraphenylporphyrinate [Bi(TPP)] as a potent protective agent against CDDP-induced nephrotoxicity. Bi(TPP) attenuates CDDP-induced acute kidney injury and prevents the death of mice exposed to a lethal dose of CDDP. The protective potency of bismuth porphyrin complexes could be optimized by varying lipophilic TPP ligands with ideal ClogP values of 8-14. Unexpectedly, Bi(TPP) exhibited a protective role via metallothionein-independent pathways, i.e., maintenance of redox homeostasis and energy supplement, elimination of accumulated platinum in the kidney, and inactivation of caspases cascade in apoptotic pathway. Significantly, Bi(TPP) does not compromise the antitumor activity of CDDP in the orthotopic tumor xenograft mouse model. These findings suggest that Bi(TPP) could be incorporated into current CDDP-based cancer therapy as a nephroprotective agent.

An Area-Efficient and Highly Linear Reconfigurable Continuous-Time Filter for Biomedical Sensor Applications.

This paper proposes a compact, high-linearity, and reconfigurable continuous-time filter with a wide frequency-tuning capability for biopotential conditioning. It uses an active filter topology and a new operational-transconductance-amplifier (OTA)-based current-steering (CS) integrator. Consequently, a large time constant τ , good linearity, and linear bandwidth tuning could be achieved in the presented filter with a small silicon area. The proposed filter has a reconfigurable structure that can be operated as a low-pass filter (LPF) or a notch filter (NF) for different purposes. Based on the novel topology, the filter can be readily implemented monolithically and a prototype circuit was fabricated in the 0.18 µm standard complementary-metal-oxide-semiconductor (CMOS) process. It occupied a small area of 0.068 mm2 and consumed 25 µW from a 1.8 V supply. Measurement results show that the cutoff frequency of the LPF could be linearly tuned from 0.05 Hz to 300 Hz and the total-harmonic-distortion (THD) was less than -76 dB for a 2 Hz, 200 mVpp sine input. The input-referred noises were 5.5 µVrms and 6.4 µVrms for the LPF and NF, respectively. A comparison with conventional designs reveals that the proposed design achieved the lowest harmonic distortion and smallest on-chip capacitor. Moreover, its ultra-low cutoff frequency and relatively linear frequency tuning capability make it an attractive solution as an analog front-end for biopotential acquisitions.

A Stochastic Quasi-Newton Method for Large-Scale Nonconvex Optimization With Applications.

Ensuring the positive definiteness and avoiding ill conditioning of the Hessian update in the stochastic Broyden-Fletcher-Goldfarb-Shanno (BFGS) method are significant in solving nonconvex problems. This article proposes a novel stochastic version of a damped and regularized BFGS method for addressing the above problems. While the proposed regularized strategy helps to prevent the BFGS matrix from being close to singularity, the new damped parameter further ensures the positivity of the product of correction pairs. To alleviate the computational cost of the stochastic limited memory BFGS (LBFGS) updates and to improve its robustness, the curvature information is updated using the averaged iterate at spaced intervals. The effectiveness of the proposed method is evaluated through the logistic regression and Bayesian logistic regression problems in machine learning. Numerical experiments are conducted by using both synthetic data set and several real data sets. The results show that the proposed method generally outperforms the stochastic damped LBFGS (SdLBFGS) method. In particular, for problems with small sample sizes, our method has shown superior performance and is capable of mitigating ill-conditioned problems. Furthermore, our method is more robust to the variations of the batch size and memory size than the SdLBFGS method.

A Novel Synthetic Compound, Bismuth Zinc Citrate, Could Potentially Reduce Cisplatin-Induced Toxicity Without Compromising the Anticancer Effect Through Enhanced Expression of Antioxidant Protein.

Cisplatin is a common anticancer drug, but it comes with significant nephrotoxicity. Further cisplatin-induced oxidative stress contributes to the pathogenesis of the nephrotoxicity. A new compound, BiZn, can potentially prevent this complication. We verified our postulation by in vitro and in vivo models. From our findings, BiZn did not affect cisplatin-induced cytotoxicity on neuroblastoma cells under both in vitro and in vivo settings. However, BiZn significantly reduced the blood urea nitrogen and creatinine levels in cisplatin-treated mice. Under the lethal dosage of cisplatin, co-treatment of BiZn significantly increased the survival rate. BiZn stimulated antioxidant proteins metallothionein (MT) and glutathione (GSH) generation from kidney cells and minimized cisplatin-induced apoptosis. Knocking down MT-IIA and inhibiting GSH abolished such protection. In conclusion, pretreatment of BiZn decreased cisplatin-induced renal toxicity without affecting its antitumor activity. BiZn-induced antioxidant proteins MT and GSH may contribute to the renal protection effect.

Purification of lysozyme from chicken egg white using diatom frustules.

A nontoxic chromatographic matrix, with low cost and high adsorption capacity, is always a major goal for therapeutic protein purification. In this study, the frustules from two cultured diatoms, Nitzschia bilobata (AQ1) and Psammodictyon panduriforme (NP), were investigated as cation exchange materials for lysozyme purification from chicken egg white. The surface area and cation exchange capacity of frustules were about 400 m2/g and 140 µmol/mL for AQ1, 390 m2/g and 130 µmol/mL for NP. The optimal pH was 9 for adsorption. Through batch purification, the lysozyme recovery was 86% with a purity of 95% by AQ1 frustules, which was higher than that by NP frustules (82% with a purity of 90%). In the flow-through system, the purity using AQ1 frustules notably increased to 99%, higher than the result of 91% using NP frustules. Diatom frustules from AQ1 are more effective and could be an alternative chromatographic matrix for lysozyme purification.

Automatic Muscle Fiber Orientation Tracking in Ultrasound Images Using a New Adaptive Fading Bayesian Kalman Smoother.

This paper proposes a new algorithm for automatic estimation of muscle fiber orientation (MFO) in musculoskeletal ultrasound images, which is commonly used for both diagnosis and rehabilitation assessment of patients. The algorithm is based on a novel adaptive fading Bayesian Kalman filter (AF-BKF) and an automatic region of interest (ROI) extraction method. The ROI is first enhanced by the Gabor filter (GF) and extracted automatically using the revoting constrained Radon transform (RCRT) approach. The dominant MFO in the ROI is then detected by the RT and tracked by the proposed AF-BKF, which employs simplified Gaussian mixtures to approximate the non-Gaussian state densities and a new adaptive fading method to update the mixture parameters. An AF-BK smoother (AF-BKS) is also proposed by extending the AF-BKF using the concept of Rauch-Tung-Striebel smoother for further smoothing the fascicle orientations. The experimental results and comparisons show that: 1) the maximum segmentation error of the proposed RCRT is below nine pixels, which is sufficiently small for MFO tracking; 2) the accuracy of MFO gauged by RT in the ROI enhanced by the GF is comparable to that of using multiscale vessel enhancement filter-based method and better than those of local RT and revoting Hough transform approaches; and 3) the proposed AF-BKS algorithm outperforms the other tested approaches and achieves a performance close to those obtained by experienced operators (the overall covariance obtained by the AF-BKS is 3.19, which is rather close to that of the operators, 2.86). It, thus, serves as a valuable tool for automatic estimation of fascicle orientations and possibly for other applications in musculoskeletal ultrasound images.