MAS509 Biomedical Signal Processing and Instrumentation (Autumn 2026)

On successful completion of the course, the student should be able to:

• Signal processing, applied AI in biomedical in imaging, and AI assisted bio-medical robotics actuation systems. Students will acquire and preprocess bio-signals, extract discriminative features, and implement ML models that translate physiological activity into control commands to actuate motorized/robotic platform targeting assistive, rehabilitation, vital sign-controlled actuation and end-to-end human-machine interfaces in general.
• Leverage AI algorithms and models such as CNN, LSTM to process bio-signals of different modality for downstream tasks such as - disease classification and prediction; anomaly detection; bio-signal and images classification, regression, and segmentation.
• Fuse AI, bio-signals and medical robotic actuation systems.
• Understand the basis in the theory of biomedical measurement systems, including sensors, signal conditioning methods, measurement techniques, patient interfacing and instrumentation used in assistive and healthcare technology
• Biomedical computer vision imaging and image analysis - Optical coherence tomography (OCT), Spectroscopy, electroencephalogram (EEG), electrocardiogram (ECG), electromyogram (EMG).
• Biomedical sensors and measurement methods.
• Signal acquisition, signal processing, fusion and feature extraction.

This course focuses on cutting-edge methods for acquiring, processing, and interpreting biomedical data. Students will learn how to acquire, preprocess, and decode bio-signals (e.g., EMG, EEG, ECG) to generate control commands for real-time actuation systems. It includes entire pipeline: feature extraction, model training, and inference that translate physiological patterns into control outputs for robotic or motorized platforms.

It introduces students to various imaging and signal modalities, such as biopotentials, electrocardiography (ECG), optical coherence tomography (OCT), spectroscopy, electroencephalography (EEG), and. The curriculum emphasizes both foundational knowledge and hands-on experience, preparing students to address real-world biomedical challenges using advanced computational techniques.

Starting with introduction of biopotentials, biomedical signal acquisition, filtering, and time-frequency analysis of bio-signals to integrating machine learning (ML) techniques into biomedical applications. Students will learn how ML can perform supervised and unsupervised learning for bio-image/signal classification, regression and clustering for downstream tasks such as disease prediction from multi-modal data, bio-signal and AI reinforced robotic actuation system. By the end of the course, participants will have a strong understanding of the theoretical concepts, practical tools, and interdisciplinary approaches required for a career in AI reinforced biomedical imaging and signal processing. This module includes the following content:

• Machine Learning for Biomedical Applications
  • Application of supervised and unsupervised ML algorithms for classification, regression, and clustering in biomedical data including OCT image, spectrometer data and other modalities for real world applications such as - disease classification and anomaly detection; bio-signal and AI reinforced robotic actuations
  • Transfer learning and custom AI/ML models leveraging Convolutional Neural Networks (CNNs) for image classification, and segmentation. Examples include - ML based retina health detection based on OCT image, Heart health detection based on ECG time series data etc.
• Biomedical computer vision imaging and image modalities analysis
  • Overview and principle of OCT, spectroscopy, EEG, ECG, and other imaging techniques used in biomedical research and clinical diagnostics.
• Bio-signal acquisition and preprocessing
  • Biomedical sensors and measurement methods
  • Techniques for bio-signal data acquisition, noise reduction, and filtering in biomedical signals
  • Fourier analysis, and wavelet transformation to process bio-signals
• Bio-signal based actuation
  • Actuate a remote robot/manipulator/motor using bio-signal towards human-machine interface, vital sign-based actuation, telemetry and rehabilitation.