Exploiting independence and causality in biomedical signal processing
Independent Component Analysis (ICA) is a technique for resolving additively mixed independent signals into their original sources. One main application of ICA is to expose the underlying sources from multiple measurements of biomedical signals such as EEG and fMRI. Traditionally ICA is performed under an assumption that the signals are real-valued and uncorrelated in time. One part of this project aims to develop algorithms that exploit the correlation structure of autocorrelated source signals. We will focus on the situation where signal ranges fall into finite alphabets. Another part of the project will examine a new notion of causal coherence and evaluate its use in biomedical applications for example in relating arterial blood pressure waveforms to intracranial pressure waveforms.

I’m a course 6-2 and 18 junior passionate in mathematics. I became interested in this project as Prof. Verghese presented to me how signal processing can be applied to healthcare problems and biomedical signals. This project will be a great opportunity for me to explore the intersection between mathematics signal processing and biomedicine.