Swarm Contrastive Decomposition
A Python/PyTorch package using particle-swarm-optimised source separation to decompose high-density EMG.
Agnese Grison
Research Associate · Machine learning for neural interfaces
Imperial–Meta Wearable Neural Interfaces Research Centre · Imperial College London
About
I develop machine-learning and signal-processing methods to extract individual motor neuron activity from high-density electromyography (EMG). I apply this work to neural interfaces and assistive technology, including research after spinal cord injury.
Since 2025, I have been a Research Associate at the Imperial–Meta Wearable Neural Interfaces Research Centre at Imperial College London, working with Prof. Dario Farina. My current work connects EMG signal processing with wearable neural interfaces and their potential applications in neurorehabilitation.
During my PhD in AI and Machine Learning in the CDT in AI4Health at the Department of Computing, Imperial College London, I studied high-density electrode arrays and methods for connecting with the human spinal cord. I also build open-source tools for EMG decomposition and review.
Research & selected work
Open-source tools
SCD Studio
A Qt/Python desktop application for GPU-accelerated decomposition and motor unit review and editing.
Motor unit decomposition
Developing methods to separate individual motor unit activity from complex EMG recordings.
Quasi-linear ICA for motor unit decomposition during dynamic contractions
NeurIPS 2026 (accepted); arXiv:2406.19581
Unlocking the full potential of high-density surface EMG: novel non-invasive high-yield motor unit decomposition
The Journal of Physiology, 603(8), 2281–2300 (2025)
A particle swarm optimized independence estimator for blind source separation of neurophysiological time series
IEEE Transactions on Biomedical Engineering, 72(1), 227–237 (2025)
Clinical interfacing
Using motor neuron activity to enable multidimensional control after spinal cord injury, including a soft assistive glove.
Intuitive, multidimensional motor unit control after paralysing spinal cord injury using non-invasive recordings
Nature Human Behaviour (accepted); medRxiv preprint, 2025
Multidimensional motoneuron control using intramuscular microelectrode arrays in tetraplegic spinal cord injury
medRxiv preprint, 2025
Hardware-informed neural interfaces
Studying how electrode design and sampling affect the motor units that can be recorded.
Motor unit sampling from intramuscular micro-electrode array recordings
IEEE Transactions on Neural Systems and Rehabilitation Engineering, 33, 620–629 (2025)
Design, fabrication and evaluation of a stretchable high-density electromyography array
Sensors, 24(6), 1810 (2024)
Neural decoding and benchmarking
Evaluating what motor neuron recordings can decode and how decomposition methods are benchmarked.
Intramuscular microelectrode arrays enable highly accurate neural decoding of finger tasks
Interface Focus, 16, 20250063 (2026)
MUniverse: A simulation and benchmarking suite for motor unit decomposition
NeurIPS 2025 · Datasets & Benchmarks Track
* Equal contribution. Selected papers; view all publications on Google Scholar.
Background
Experience
- 2025–Present
- Research Associate, Imperial–Meta
Wearable Neural Interfaces Research Centre, Imperial College London
Supervisor: Prof. Dario Farina - 2020–2025
- Graduate Teaching Assistant, Imperial College London
Mathematics I & II, Digital Biosignal Processing for Human Monitoring, and Brain–Machine Interfaces - Jan 2024
- Visiting Student, Institute of Neuroinformatics, University of Zurich and ETH Zurich
- 2021
- Data Scientist, Mindstep
Developed a cognitive score and analysed app data to support screening for early signs of dementia. - 2020
- Research Assistant, Brain and Vestibular Group, Imperial College London
Studied movement, balance, and orientation using clinical data and neurophysiological recordings. - 2018–2020
- Undergraduate Teaching Assistant, Imperial College London
Image Processing; Mathematical Tools, Vibrations and Waves - 2018–2020
- Sales Assistant, Imperial College Union
- 2018
- Undergraduate Research Opportunities Programme intern, Imperial College London
Developed an app for a blood vancomycin biosensor with a multidisciplinary team for the SensUs competition. - 2015
- Distinguished Visiting Scholar, Houston Methodist Research Institute
Shadowed researchers.
Education
- 2021–2025
- PhD in AI and Machine Learning, CDT in AI4Health, Department of Computing, Imperial College London
Supervisor: Prof. Dario Farina
Thesis: Implanted high-density electrode arrays to connect with the human spinal cord - 2016–2020
- MEng in Biomedical Engineering, Imperial College London
Contact
For questions or collaborations in motor unit decomposition, neural interfaces, and assistive technology, email agnese.grison@outlook.it.
GitHub · Google Scholar · ORCID · LinkedIn
Outside the lab
Making things
I grew up in Udine and joined the AGESCI Udine 1 Scout group when I was seven. One of my earliest memories there is earning the “Mani abili” (“Handicraft”) badge. Making things with my hands has always mattered to me; it was one of the main reasons I chose engineering. I still try to make time for it, whether I'm knitting, painting, or making pasta from scratch (the Italian in me). The tortellini here come in two versions: one crocheted, one made from dough.
Philosophy
My dad was a philosophy teacher, so philosophy was part of my life long before I studied it in high school. I've loved it ever since. During those years, I explored the work of Simone Weil, P. A. Florenskij, Hannah Arendt, Emmanuel Levinas, and F. M. Dostoevskij in depth.
When I think about philosophy, I picture Aristotle and his students discussing ideas as they walked. Some of my own best ideas come on a walk, and I hope to keep finding time to walk and think for as long as I can.
Outdoors
I've always enjoyed being outdoors and staying active. When I was little, climbing meant reaching the top of our bookcase and nearly giving my grandmother a heart attack; these days, I prefer hiking and bouldering. I love being on the water too - it reminds me of Trieste, a city I love. I joined the Imperial College Boat Club in my first year at Imperial. It was hard work, but I loved every minute of it.
