All programmes > Brain Computer Interfaces

Brain Computer Interfaces (B-KUL-H08M0A)

4 ECTSEnglish31 Second term
Van Hulle Marc
POC Artificial Intelligence

Aims

Brain Computer Interfaces (BCIs) are aimed at creating a direct communication pathway between the brain and an external device, bypassing the need for an embodiment. Research in this field has witnessed a spectacular development, and BCIs are nowadays regarded as one of the most successful engineering applications of the neurosciences. Indeed, such systems can provide a significant improvement of the quality of life of neurologically impaired patients suffering from amyotrophic lateral sclerosis (ALS), stroke, brain/spinal cord injury, muscular dystrophy, etc. In addition, it also been used in communication-, motor revalidation-, motor substitution- and entertainment applications (gaming).

In this course, first basic knowledge of the anatomy and physiology of the brain is given, and of the type of signals that are recorded for BCI purposes. Then, the invasive BCIs are discussed, what type of signal features are extracted, and how classifiers and regressors are built. Several case studies are discussed: text spelling, robot arm and exoskeleton control, speech and handwriting decoding.

Then, the noninvasive BCIs are discussed, thereby mostly concentrating on the EEG-based ones. Several case studies are introduced, involving text spelling, semantics, emotion detection,...

Previous knowledge

Basic knowledge of signal processing.

Is included in these courses of study

Activities

3.5 ects. Brain Computer Interfaces: Lectures (B-KUL-H08M0a)

3.5 ECTSEnglishFormat: Lecture21 Second term
Van Hulle Marc
POC Artificial Intelligence

Content

1. Introduction

Definition of Brain Computer Interfaces.

Types of interfaces (invasive and noninvasive), developments and testing, applications.

2. Basic principles of Neuroscience

Anatomy and physiology of human and monkey brain.

Brain signals: spikes, local field potentials (LFPs), electrocorticography (ECoG), EEG, fMRI,...

3. Invasive BCI

Definition, type of signals (spikes, LFPs, ECoGs), recording methodology, recording sites, signal conditioning, feature construction, feature selection, decoding (classification/regression). Examples of invasive BCIs: text spelling, decoding and tracking arm (hand) position, controlling prosthetic devices such as orthotic hands, robot arms and exeskeletons, speech and handwrtiing decoding.

4. Noninvasive BCI

Definition, type of signals (EEG, fMRI), comparison with invasive BCI (lower spatial and/or temporal resolution), signal conditioning, feature construction, feature selection, decoding (classification/regression).

Examples of noninvasive BCIs based on visually evoked potentials (VEPs) mu-rhythms, event-related potentials (ERPs).

Course material

Course material downloadable from Toledo.

Language of instruction: more information

English

Format: more information

Regular ex-cathedra teaching with case studies and examples to promote student interaction (questions).

0.5 ects. Brain Computer Interfaces: Exercises (B-KUL-H08M1a)

0.5 ECTSEnglishFormat: Practical10 Second term
Van Hulle Marc
POC Artificial Intelligence

Content

2 lab sessions are planned during which the student gets hand-on experience with EEG-based BCI.

Course material

Matlab code provided during lab session.

Language of instruction: more information

English.

Format: more information

The student will perform an EEG experiment and analyse the results using available Matlab code.

Evaluation

Evaluation: Brain Computer Interfaces (B-KUL-H28M0a)

Type : Exam during the examination period
Description of evaluation : Written
Type of questions : Closed questions
Learning material : None

Explanation


Written exam. Example questions are available from the course's Toledo page.