Research Assistant: EEG-based Brain-Computer Interface for Motor Imagery

Location: Stockholm

Type: Research

Term: Semester

Introduction to the topic

Can we learn physical skills without moving a muscle? Motor imagery—the act of vividly imagining a movement—activates the same neural pathways as physical execution. This project aims to build the foundation for Momente Neurotechnologies, a non-invasive brain-computer interface (BCI) startup designed to help musicians and athletes practice mentally.

This semester focuses on the build phase. The goal is to design, code, and validate a functional electroencephalography-based BCI prototype that is robust enough for future human testing.

We are looking for three Research Assistants to form an interdisciplinary team.

The neural engineer: Analyze EEG data with signal processing and machine/deep learning techniques. The neural engineer will gain hands-on experience with biosignal processing (filtering EEG noise, features extraction), machine and deep learning classifiers (LDA/SVM/EEGNet/transformer-based models), and Python programming for analyzing scientific data.
The product developer: Create the visual feedback application (game development, UX design). The product developer will gain experience with real-time application architecture, integrating Python with UI frameworks, and designing user-centric biofeedback interfaces.
The protocol lead: Design experimental protocols, validate neuroscience aspects, and write ethics approvals for human experiments. The protocol lead will gain practical skills in experimental design, creating standard operating procedures (SOP), developing calibration protocols, creating motor imagery instructions, understanding of EEG neuroanatomy, and navigating research ethics applications.

If you are a student who wants to move beyond theory and get hands-on experience building a real neurotech system, this project is for you.

Project details

BCI systems are often fragile and confined to research laboratories. To bring them into real-world settings, there is a need for robust and user-friendly systems. In the context of non-invasive EEG-based BCI systems, motor imagery is a particularly relevant paradigm, as it can be used both to study motor-related brain activity and to provide neurofeedback for training or rehabilitation.

This project focuses on the development of an EEG-based motor imagery BCI platform that connects raw EEG acquisition to real-time signal processing, classification, and visual feedback. Beyond the technical challenge of detecting motor imagery reliably, the project is motivated by two research questions:

Can users learn to self-modulate an EEG-derived motor imagery neurofeedback target?
Does this self-modulation have the potential to transfer to improved motor performance?
The objective of this semester is to deliver a foundational, deployment-ready BCI software platform. By the end of the semester, the team will assemble a software stack connecting raw EEG to a visual feedback application, validate the algorithmic pipeline through internal self-experimentation, and draft standard operating procedures (SOPs) for future data collection. This semester will therefore lay the technical and methodological groundwork for later experimental studies involving external participants and task-specific motor performance outcomes.

Methodology
This project operates as a research sprint utilizing self-experimentation and iterative design.

Hardware/software loop: research assistants will write code iteratively and test it on their own brains to identify bugs or latency.
Protocol development: The team will run internal alpha tests to refine user instructions and interface rather than recruiting external subjects immediately.
Literature review and documentation: A critical component is documenting the code, experimental procedures, and design choices for future continuity. The documentation will also include a literature review on EEG-based motor imagery.

Research note: This is a foundational semester. Students will most likely not be involved in large-scale external data collection during this phase. Instead, they will build the software, documentation, and experimental procedure required for future studies, laying the groundwork that makes such research possible.

Syllabus

Coming soon

Selected relevant publications:

– Stefano Filho, C.A., Attux, R. & Castellano, G. (2024). Motor Imagery Neurofeedback: From System Conceptualization to Neural Correlates. Curr Behav Neurosci Rep 11, 78–98.
– Lawhern, Vernon & Solon, Amelia & Waytowich, Nicholas & Gordon, Stephen & Hung, Chou & Lance, Brent. (2018). EEGNet: a compact convolutional neural network for EEG-based brain–computer interfaces. Journal of Neural Engineering. 15. 10.1088/1741-2552/aace8c.
– Tidare, J., Leon, M., & Astrand, E. (2021). Time-resolved estimation of strength of motor imagery representation by multivariate EEG decoding. Journal of neural engineering, 18(1), 10.1088/1741-2552/abd007.

Prerequisites

No prior experience with BCI is required, but a strong willingness to learn and experiment is essential.

• For the neural engineer: Proficiency in Python (NumPy, PyTorch), comfortable with mathematics (linear algebra, probability theory, statistics, optimization), and great coding skills are expected.

• For the product developer: Experience in interactive application development (UI/UX) or game development.

• For the protocol lead: Scientific literacy and basic knowledge of neuroanatomy and experimental design. Students are not expected to know neuroscience/neuropsychology in depth, but they should be able to read scientific papers, recognize what they do not yet understand, and independently search for the information needed to understand motor imagery and neurofeedback protocols.

Additional research application required

You will need to submit an additional research application through Student Registration in order to enroll in this course.
To submit your research application, you must already be admitted to DIS.

All research application materials must be submitted on the following dates by 23:59 Central European Time:
– November 1 for spring semester applicants
– May 1 for fall semester applicants

Complete your additional research application through Student Registration.
If you are not already enrolled, use this link to apply to DIS before completing a research application.

Aniss Aiman Medbouhi

Ph.D. candidate in Computer Science, KTH Royal Institute of Technology (2022–present). M.Sc. double degree in Computer Science and Engineering/General Engineering, KTH Royal Institute of Technology/Ecole Centrale Marseille (2017–2021). Previous experience developing an algorithm of toxicity risk prediction for anti-cancer chemotherapy and supporting KTH’s Machine Learning and Artificial Intelligence courses. With DIS since 2025.

Research Assistantship FAQ

Learn more about research at DIS.

: Earning authorship within the course of a semester is challenging. Exceptional Research Assistants may be considered for co-authorship, subject to evaluation from the supervising Research Mentor.

Though you may not ultimately become an author on a published paper, you will gain many practical skills and insights into the publication process.

: Many research projects last across several semesters, so not all RAs on the same project will conduct the same research.

: On average, students in 3-credit Research Assistantships spend at least 10 hours a week on research tasks. These 10 hours per week consist of a variety of different research-related tasks, including conducting a literature review, data collection, analyses, coding, documentation, and meetings with your mentor. You may decide on a particular block of time (consistent each week) for you to carry out the work, or adjust it on a week-by-week basis.

Students in a 6-credit Research Assistantships spend an average of 20 hours per week with lab and clinically based research tasks. You decide the exact meeting and lab times with your supervisor, but your schedule consists of classes at DIS two days a week, with two days a week in the lab.

There may be peak times in the research process where all Research Assistant students are expected to spend a few more hours, and then possibly fewer hours another week – to reflect the individual research project and process.

: Work for the 3-credit Research Assistantships may take place outside DIS, on your own time at a location of your choosing, or at DIS – this varies from project to project. The 6-credit research assistantships generally take place at labs in Copenhagen or Stockholm outside DIS.

: Students with a variety of academic backgrounds are encouraged to apply for the 3-credit Research Assistant course. While some courses do have prerequisites, many opportunities do not require prior research experience.

The 6-credit Research Assistant opportunity is designed for highly motivated students who have a strong science background within the field of interest and some basic research experience. The student must be willing to commit a large portion of their semester to an engaging opportunity involving hands-on research in universities and medical institutes in Copenhagen or Stockholm.