Taipan: Causal Streaming State-Space Decoding for Real-Time fMRI
About
Real-time fMRI decoding could enable closed-loop neurofeedback, but most decoders rely on offline data unavailable during live scans. Here, Phillip presents "Taipan", a causal streaming state-space model that decodes cognitive states one TR at a time. Taipan matches or outperforms offline GLM baselines across two independent datasets - decoding sustained valence in prospective imagery and transferring zero-shot to unseen subjects in intertemporal choice. To bridge to live scanner sessions, ongoing work demonstrates that pre-session structural registration maintains full decoding performance under strict online causality constraints.
Details
- Date:
- Friday, October 30, 2026
- Time:
- 12:00 PM - 1:00 PM
- Location:
- CDS 1646
- Theme:
- Computational Methodology
Speaker
Phillip Angelos
Phillip Angelos is a PhD student in the Faculty of Computing and Data Sciences at Boston University, advised by Dr. Joshua Peterson and Dr. Arthur Lee. He earned a Bachelor of Science in Psychology from Michigan State University and spent two years at Yale University researching positive symptom progression in psychosis. His research examines the intersection of artificial intelligence and psychology, with a focus on deep learning, decision-making, impulsivity, and related behavioral patterns.