Towards a wearable low-cost ultrasound device for classification of muscle activity and muscle fatigue

Brausch, Lukas; Hewener, Holger; Lukowicz, Paul

doi:10.1145/3341163.3347749

Cited by 11 publications

(5 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Figure 1 illustrates the examined muscles or muscle groups with the gastrocnemius muscle sketched on the left and the biceps brachii muscle sketched on the right. This work builds upon and extends previously published preliminary results [ 22 , 23 ].…”

Section: Introductionsupporting

confidence: 84%

“…For the muscle contraction data classification, we included the ML models MLP, FCN, ResNet, ROCKET, MINIROCKET, MultiRocket, CatBoost, XGBoost, LightGBM, Transformer, 1-NN DTW, SVM and Logistic Regression. For the muscle fatigue data classification, we omitted the models MLP, FCN, and ResNet as their inclusion would have led to a massive increase in computation time by several months for each model, while the expected performance improvement, judging from the results obtained for the muscle contraction classifications, was low [ 22 ]. This approach led to 252 possible combinations for the muscle contraction data and 2376 possible combinations for the muscle fatigue data.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Classifying Muscle States with One-Dimensional Radio-Frequency Signals from Single Element Ultrasound Transducers

Brausch

Hewener

Lukowicz

2022

Sensors

Self Cite

View full text Add to dashboard Cite

The reliable assessment of muscle states, such as contracted muscles vs. non-contracted muscles or relaxed muscles vs. fatigue muscles, is crucial in many sports and rehabilitation scenarios, such as the assessment of therapeutic measures. The goal of this work was to deploy machine learning (ML) models based on one-dimensional (1-D) sonomyography (SMG) signals to facilitate low-cost and wearable ultrasound devices. One-dimensional SMG is a non-invasive technique using 1-D ultrasound radio-frequency signals to measure muscle states and has the advantage of being able to acquire information from deep soft tissue layers. To mimic real-life scenarios, we did not emphasize the acquisition of particularly distinct signals. The ML models exploited muscle contraction signals of eight volunteers and muscle fatigue signals of 21 volunteers. We evaluated them with different schemes on a variety of data types, such as unprocessed or processed raw signals and found that comparatively simple ML models, such as Support Vector Machines or Logistic Regression, yielded the best performance w.r.t. accuracy and evaluation time. We conclude that our framework for muscle contraction and muscle fatigue classifications is very well-suited to facilitate low-cost and wearable devices based on ML models using 1-D SMG.

show abstract

Section: Introductionsupporting

confidence: 84%

Section: Methodsmentioning

confidence: 99%

Classifying Muscle States with One-Dimensional Radio-Frequency Signals from Single Element Ultrasound Transducers

Brausch

Hewener

Lukowicz

2022

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…Quantitative ultrasound biomarkers derived from raw data have consistently demonstrated their ability to capture both tissue composition and microstructural properties. Furthermore, they have exhibited a greater capacity to encode richer information content compared with ultrasound B-mode images in artificial intelligence models, as highlighted in prior research [52,54,83,84]. Ultrasound spectroscopy parameters and tissue acoustic properties, such as speed of sound and attenuation, have been associated with tissue composition [47] and viscoelastic changes in muscle [49] among older individuals with sarcopenia.…”

Section: Comparison With Prior Workmentioning

confidence: 94%

“…Textural radiomics [52,[82][83][84] Radiofrequency data and B-mode features extracted automatically with end-to-end neural network models trained with respect to clinical outcomes…”

Section: Automatic Morphometric Measurementsmentioning

confidence: 99%

Development of Continuous Assessment of Muscle Quality and Frailty in Older Patients Using Multiparametric Combinations of Ultrasound and Blood Biomarkers: Protocol for the ECOFRAIL Study

Virto,

Río,

Angulo-Garay

et al. 2024

JMIR Res Protoc

View full text Add to dashboard Cite

Background Frailty resulting from the loss of muscle quality can potentially be delayed through early detection and physical exercise interventions. There is a demand for cost-effective tools for the objective evaluation of muscle quality, in both cross-sectional and longitudinal assessments. Literature suggests that quantitative analysis of ultrasound data captures morphometric, compositional, and microstructural muscle properties, while biological assays derived from blood samples are associated with functional information. Objective This study aims to assess multiparametric combinations of ultrasound and blood-based biomarkers to offer a cross-sectional evaluation of the patient frailty phenotype and to track changes in muscle quality associated with supervised exercise programs. Methods This prospective observational multicenter study will include patients aged 70 years and older who are capable of providing informed consent. We aim to recruit 100 patients from hospital environments and 100 from primary care facilities. Each patient will undergo at least two examinations (baseline and follow-up), totaling a minimum of 400 examinations. In hospital environments, 50 patients will be measured before/after a 16-week individualized and supervised exercise program, while another 50 patients will be followed up after the same period without intervention. Primary care patients will undergo a 1-year follow-up evaluation. The primary objective is to compare cross-sectional evaluations of physical performance, functional capacity, body composition, and derived scales of sarcopenia and frailty with biomarker combinations obtained from muscle ultrasound and blood-based assays. We will analyze ultrasound raw data obtained with a point-of-care device, along with a set of biomarkers previously associated with frailty, using quantitative real-time polymerase chain reaction and enzyme-linked immunosorbent assay. Additionally, we will examine the sensitivity of these biomarkers to detect short-term muscle quality changes and functional improvement after a supervised exercise intervention compared with usual care. Results At the time of manuscript submission, the enrollment of volunteers is ongoing. Recruitment started on March 1, 2022, and ends on June 30, 2024. Conclusions The outlined study protocol will integrate portable technologies, using quantitative muscle ultrasound and blood biomarkers, to facilitate an objective cross-sectional assessment of muscle quality in both hospital and primary care settings. The primary objective is to generate data that can be used to explore associations between biomarker combinations and the cross-sectional clinical assessment of frailty and sarcopenia. Additionally, the study aims to investigate musculoskeletal changes following multicomponent physical exercise programs. Trial Registration ClinicalTrials.gov NCT05294757; https://clinicaltrials.gov/ct2/show/NCT05294757 International Registered Report Identifier (IRRID) DERR1-10.2196/50325

show abstract

“…Advances would also be provided by practical interfaces, for example wearable interfaces that can be easily donned such as EMG bracelets [180]. Other techniques to acquire signals stemming from muscle activities, such as wearable ultrasound devices, are being developed [181].…”

Section: Limitations and Future Directionsmentioning

confidence: 99%

A review on interaction control for contact robots through intent detection

Sena

Wang

et al. 2022

Prog. Biomed. Eng.

View full text Add to dashboard Cite

Interaction control can take opportunities offered by contact robots physically interacting with their human user, such as assistance targeted to each human user, communication of goals to enable effective teamwork, and task-directed motion resistance in physical training and rehabilitation contexts. Here we review the burgeoning field of interaction control in the control theory and machine learning communities, by analysing the exchange of haptic information between the robot and its human user, and how they share the task effort. We first review the estimation and learning methods to predict the human user intent with the large uncertainty, variability and noise and limited observation of human motion. Based on this motion intent core, typical interaction control strategies are described using a homotopy of shared control parameters. Recent methods of haptic communication and game theory are then presented to consider the co-adaptation of human and robot control and yield versatile interactive control as observed between humans. Finally, the limitations of the presented state of the art are discussed and directions for future research are outlined.

show abstract

Towards a wearable low-cost ultrasound device for classification of muscle activity and muscle fatigue

Cited by 11 publications

References 6 publications

Classifying Muscle States with One-Dimensional Radio-Frequency Signals from Single Element Ultrasound Transducers

Classifying Muscle States with One-Dimensional Radio-Frequency Signals from Single Element Ultrasound Transducers

Development of Continuous Assessment of Muscle Quality and Frailty in Older Patients Using Multiparametric Combinations of Ultrasound and Blood Biomarkers: Protocol for the ECOFRAIL Study

A review on interaction control for contact robots through intent detection

Contact Info

Product

Resources

About