A Semiotic Framework for Multimodal Fusion in Stroke Prognosis

This PhD position at Université Paris-Saclay's IBISC Laboratory focuses on advancing multimodal data fusion by incorporating a pragmatic dimension alongside traditional syntactic and semantic approaches. The research integrates semiotic theory and a Partial Information Decomposition framework to process complex medical data for stroke prognosis. The successful candidate will collaborate closely with an interdisciplinary team of data scientists and clinicians from partner hospitals. The work expects strong competencies in adaptive learning algorithms and Python programming within a healthcare-focused engineering environment.

• Design and develop adaptive learning algorithms for processing and indexing complex medical data.
• Formulate and operationalize a semiotic framework for multimodal data fusion across syntactic, semantic, and pragmatic levels.
• Implement Partial Information Decomposition (PID) models to categorize fusion strategies based on redundancy, complementarity, and synergy.
• Integrate multimodal algorithms into operational systems in collaboration with clinical partners and data scientists.
• Engage in interdisciplinary research bridging machine learning, neuroscience, and medical applications.