Estimating Force Exerted by the Fingers Based on Forearm Ultrasound

Biosignal-based hand movement and force assessment are critical for human-machine interaction. Surface electromyography has been used to predict finger forces, but there are issues surrounding the sensor signal-to-noise ratio, number of sensors required to get good data, etc. [1-2]. Ultrasound can be used to visualize and analyze the forearm cross-section to estimate hand movements [3-4]. Recent work has shown that ultrasound can be used to estimate isometric hand forces [5]. Merely isometric grasp force estimation is not sufficient, and it’s important to get finer force measurement per finger to get force feedback for effective human-machine interfacing. In this work, we show that we can use Machine Learning models for binary force classification and continuous estimation of finger force using forearm ultrasound data.

PDF: Paper.

References -

1. C. Castellini, and R. Koiva. ”Using surface electromyography to predict single finger forces.” 2012 4th IEEE RAS & EMBS International Conference on Biomedical Robotics and Biomechatronics (BioRob). IEEE, 2012.
2. M. Zheng, M. S. Crouch, and M. S. Eggleston. ”Surface electromyography as a natural human–machine interface: a review.” IEEE Sensors Journal 22.10 (2022): 9198-9214.
3. K. Bimbraw, c. J. Nycz, M. J. Schueler, Z. Zhang, & H. K. Zhang. ”Prediction of Metacarpophalangeal joint angles and Classification of Hand configurations based on Ultrasound Imaging of the Forearm.” 2022 International Conference on Robotics and Automation (ICRA). IEEE, 2022.
4. K. Bimbraw, C. J. Nycz, M. J. Schueler, Z. Zhang, & H. K. Zhang. ”Simultaneous Estimation of Hand Configurations and Finger Joint Angles Using Forearm Ultrasound.” IEEE Transactions on Medical Robotics and Bionics 5.1 (2023): 120-132.
5. A. T. Kamatham, M. Alzamani, A. Dockum, S. Sikdar, & B. Mukherjee. ”SonoMyoNet: A Convolutional Neural Network for Predicting Isometric Force From Highly Sparse Ultrasound Images.” bioRxiv (2022): 2022-06.