Overview
Our research in Biomedical Systems is dedicated to next-generation, portable diagnostic and therapeutic devices that integrate advanced sensing, low-power signal processing, and AI-driven interpretation. We are pioneering the development of indigenous brain–computer interfaces alongside compact ultrasound imaging platforms—each engineered for real-time, point-of-care use and optimized for minimal data redundancy and energy consumption.
Key Research Areas
High-Density Brain–Computer Interfaces
Modular high channel count neuromorphic front ends paired with ML co-processors enable scalable, real-time neural decoding for neurorehabilitation and cognitive monitoring using invasive/non-invasive neural signals (EEG/ECoG/iEEG/EMG etc).
SUSHRUT – Super Resolution Small Handheld Reconfigurable Ultrasound Imaging System
Our research in non-linear beamforming ultrasound technology and edge AI culminates in SUSHRUT(Super Resolution Small Handheld Reconfigurable Ultrasound Imaging System)—a fully integrated hardware-software platform for portable medical imaging. Designed with a focus on accessibility, adaptability, and compliance, Sushrut embodies the future of point-of-care diagnostics. We engineer advanced beamforming algorithms, including nonlinear and pixel-based techniques, optimized for on-device execution without dependence on high-power servers. By embedding AI accelerators directly into the hardware, Sushrut enables real-time, intelligent analysis—even in resource-limited settings. With support for 2D, 3D, and Doppler imaging, along with raw data extraction and 15 MHz transducer bandwidth, Sushrut is versatile and future-proof. Its real-time reconfigurability allows seamless adaptation to various transducers, while ensuring compliance with regulations such as the PCPNDT Act through imageless detection. Available in palmtop, handheld, and portable configurations, Sushrut exemplifies how compact design, hardware-software co-design, and domain-specific AI can come together to deliver scalable, reliable, and Made-in-India medical imaging solutions.
Selected Publications
- Krishna, A., Ramanathan, V., Yadav, S. S., Shah, S., van Schaik, A., Mehendale, M., & Thakur, C. S. (2023, October). A Sparsity-driven tinyML Accelerator for Decoding Hand Kinematics in Brain-Computer Interfaces. In 2023 IEEE Biomedical Circuits and Systems Conference (BioCAS) (pp. 1–5). IEEE.
- Mahurkar, A. G., Pokala, P. K., Thakur, C. S., & Seelamantula, C. S. (2019, May). SAMIR: Sparsity Amplified Iteratively-Reweighted Beamforming for High-Resolution Ultrasound Imaging. In ICASSP 2019—IEEE International Conference on Acoustics, Speech and Signal Processing (pp. 1045–1049). IEEE.
- Gupta, S. K., Kumar, K., Seelamantula, C. S., & Thakur, C. S. (2019, July). A Portable Ultrasound Imaging System Utilizing Deep Generative Learning-Based Compressive Sensing on Pre-Beamformed RF Signals. In 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 2740–2743). IEEE.