Syllabus Application
EE 58010
Special Topics in EE: Brain-Computer Interfaces
Faculty
Faculty of Engineering and Natural Sciences
Semester
Fall 2026-2027
Course
EE 58010 -
Special Topics in EE: Brain-Computer Interfaces
Time/Place
Time
Week Day
Place
Date
11:40-13:30
Mon
FENS-G025
Sep 28-Dec 31, 2026
08:40-09:30
Wed
FENS-L065
Sep 28-Dec 31, 2026
Level of course
Masters
Course Credits
SU Credit:3, ECTS:10
Prerequisites
-
Corequisites
EE 5801L
Course Type
Lecture
Instructor(s) Information
Hüseyin Özkan
- Email: hozkan@sabanciuniv.edu
Course Information
Catalog Course Description
This course introduces EEG-based Brain-computer Interfaces (BCIs) from end to end: quality control, preprocessing (filtering, artifact handling/ICA), epoching, feature extraction, decoding, evaluation (with strong emphasis on avoiding data leakage), and decision logic for feedback/control. Students implement and evaluate at least two BCI paradigms, including P300 and SSVEP (with motor imagery as an additional paradigm), and connect design choices to latency, accuracy, and robustness in both offline and streaming settings. The course builds the necessary machine learning and signal processing tools that are commonly used in BCIs.
Learning Outcomes:
| 1. | Build a reproducible EEG pipeline: Quality control - preprocessing - epoching - features - decoding - evaluation |
|---|---|
| 2. | Implement a BCI in at least two different paradigms (P300, SSVEP, MI) |
| 3. | Evaluate BCIs without leakage, report metrics, and perform ablations |
| 4. | Apply machine learning techniques for signal processing and detection in BCIs |
| 5. | Address user-to-user and session-to-session statistical variations with domain adaptation/generalization techniques from machine learning |
| 6. | Convert offline decoding to streaming, and build a closed-loop system simulation with latency measurement |
| 7. | Explain other BCI modalities and connect them to applications and constraints |
| 8. | Demonstrate introductory-level awareness of recent developments in the BCI engineering literature |
Course Objective
This course introduces EEG signal processing for brain-computer interfaces (BCIs) from end to end: quality control, preprocessing, epoching, feature extraction, decoding, evaluation. Students implement and evaluate at least two BCI paradigms, including but not limited to motor imagery, P300 or SSVEP.
-
Course Materials
Resources:
Brain-Computer Interfaces by Bin He, Han Yuan, Jianjun Meng, and Shangkai Gao in the book: He, B. ed., 2020. Neural engineering. Springer Nature.
Rao, R.P., 2013. Brain Computer Interface: An Introduction. Cambridge University Press, New York.
Luck, S.J., 2014. An introduction to the event-related potential technique. MIT press.
Rao, R.P., 2013. Brain Computer Interface: An Introduction. Cambridge University Press, New York.
Luck, S.J., 2014. An introduction to the event-related potential technique. MIT press.
Technology Requirements:
Software: Python with (i) NumPy and SciPy for numerical computation, spectral analysis, and digital filter design; (ii) MNE-Python for EEG preprocessing, epoching, ICA, time–frequency analysis, and visualization; (iii) MOABB for standardized access to public BCI datasets and benchmarking tools; and (iv) scikit-learn for baseline machine learning such as dimensionality reduction, classification, model selection, and performance evaluation. Programming exercises use Jupyter notebooks, with Matplotlib for plotting. All laboratories use simulated signals or prerecorded EEG; no acquisition hardware is required.