Executive Summary : | Biomedical signals, such as electrocardiogram (ECG), electroencephalogram (EEG), and electromyogram (EMG), are crucial for early warning of patient diseases. These signals, which are non-invasive, provide valuable information about heart and brain abnormalities. However, noises can obscure meaningful information during measurement, necessitating the development of automated and improved diagnosis systems. Wearable signal analysis and detection systems can provide patients with immediate and continuous health feedback. However, most biomedical agencies have not developed such wearable systems due to their high costs. The objective is to design a hardware prototyping framework for processing ECG and EEG signals, indicating anomalies in the heart and brain. The main challenge is to design a fast, automated, and reliable prediction unit that can be used for general-purpose usage, especially for emergency and remote patients. The process includes designing noise reduction algorithms, designing feature extraction units, designing simple machine learning algorithms, and developing the entire prediction unit using indigenous ip-core processor (VEGA/sHAKTI)-based Field Programmable Gate Array (FPGA) prototyping. Innovations include validation of output results with ground truth delineation for reliability testing, and technology transfer of netlist design for developing a cost-effective Application specific Integrated Circuit (AsIC) in the future. |