Neural signals and signal processing

Objectives

Understanding, processing, and analysis of signals and images obtained from the central and peripheral nervous system

Course Summary

Understanding neural signals obtained by electrophysiology and imaging techniques requires knowledge both about their origin and the measurement process. This course will introduce the properties of a wide range of neural signals that are used to study the brain in health and disease. The relevance of these signals for applications in fundamental and clinical neuroscience will be made clear. In addition, a broad range of signal processing tools and their implementations will be presented with the specific focus to implement and tailor analysis of these signals, which typically comes as large, noisy, but richly structured datasets. Exercises and lab exercises will provide insights into the analysis of imaging data and electrophysiological neural signals.

Textbooks

1. H. Op de Beeck, C. Nakatani, Introduction to Human Neuroimaging,  Cambridge University Press, 2019, DOI
2. N. V. Thakor (Editor), Handbook of Neuroengineering, Springer, 2020

Assessment

The objectives of the courses will be assessed as follows

• 50% on final exam: written exam, closed book, but hand-written (not printed) cheat sheet is allowed
• 50% on two mini-projects: two recordings and reports (2x 50%)

Mini-projects will be carried out in groups. 

We strongly encourage you to attend all the the lectures and exercise sessions. These in-person activities will help you master the content of the course and prepare you for the mini-projects and final exam.

Forum

The forum offers a place for discussions among you and your classmates. If you would like to get your questions answered by the TA team, we encourage you to attend the dedicated exercise or lab sessions. In particular, if you have questions about the lectures, we encourage you to directly discuss with the professor after class.

Video-lectures

It is strongly recommended to attend lectures and exercises in person to successfully pass this course. Nevertheless, recordings of the lectures will be made on a best-effort basis and put available on Mediaspace. 

Week	Lectures (Tue 9h15-13h)	Exercises (Thu 16h15-18h)	Mini-projects
(1) 10.09 - 12.09	Part 0: Data science perspective on (f)MRI	Visualization (f)MRI
(2) 17.09 - 19.09	Part 1-1: Basics of neuroscience Part 1-2a: Basics of MRI	Preprocessing (f)MRI
(3) 24.09 - 26.09	Part 1-2b: Basics of MRI Part 1-3: Structural imaging	Preprocessing (f)MRI and dMRI
(4) 01.10 - 03.10	Part 2-1: Hemodynamic imaging Part 2-2: Experiment design Part 2-3a: GLM	Preprocessing (f)MRI and fNIRS
(5) 08.10 - 10.10	Part 2-3b: GLM Part 2-4: FC and multivariate	Task fMRI and GLM
(6) 15.10 - 17.10	Part 2-5: Connectomics + 2h help session	2h help session	Hand-out project 1
(7) 29.10 - 31.10	Part 2-5: Connectomics Part 2-6: MVPA	Multivariate approaches
(8) 5.11 - 7.11	Part 3: M/EEG + 2h help session	Functional connectivity, graphs and connectors	Deadline project 1 - Thursday 16h00
(9) 12.11 - 14.11	Part 3: M/EEG	EEG and MEG
(10) 19.11 - 21.11	Part 4a: EMG	EMG data preprocessing
(11) 26.11 - 28.11	Part 4b: Hybrid and biomechanics models	Feature engineering	Hand-out project 2
(12) 3.12 - 5.12	Part 4c: Cortical	Training and validation of ML models
(13) 10.12 - 12.12	Part 4d: Cortical/ENG	2h help session
(14) 17.12 - 19.12	Part 4e: Final overview and Q&A	Simulated recordings via hybrid modelling	Deadline project 2 - Thursday 16h00

For the links to Google Drive, use your EPFL login (not your Gmail!) to obtain access to the material.