Course information

What?
The lecture series contains of 9 lectures and two seminars at the end in which students give short presentations. We will cover fundamentals of various biosignals, timeseries pre-processing, decomposition methods, feature extraction and typical challenges in multivariate / multimodal biosignal analysis. The course is based on common methods and challenges that the Intelligent Biomedical Sensing Lab is working on towards wearable neurotechnology and brain-body imaging.

When/Where:
This lecture series will take place Tuesday 10-12:00 in-person in room MAR 4.033 starting October WS - 2025.

How to register:

Please register for the course in MOSES or via email to secure a spot. Registration in ISIS will not be considered binding.
MOSES link to the module [tbd]
MOSES link to the course [tbd]

Zoom link (online attendance is possible in justified cases)

Prerequisites:
A B.Sc. degree in computer science, biomedical or electrical engineering or a comparable field is recommended. Prior attendence of Machine Learning 1 and basic knowledge of linear algebra is expected.

Grading:

The module is pass/fail based on an online multiple choice test.

We will require students to upload their slides prior to the presentation through the "presenter slides" function.

Learning Outcomes:
Knowledge and skills: Students will understand the basic concepts of common biosignals from the domains of electrophysiology (EEG, EOG, EMG, ECG, EDA), diffuse optics (fNIRS, DOT, PPG), and a few selected others. They will be familiar with filtering techniques for timeseries signal (pre-)processing, such as frequency filters, bode diagrams, fourier analysis (nyquist theorem) and wavelet analysis. They will know different frameworks for decomposition of multivariate timeseries signals, both supervised and unsupervised linear models. Basics of sensor and feature fusion and feature extraction of timeseries signals and typical challenges (artefacts, data synchronization, explainability) will be understood.