Multimodal Data Integration and Analytics

Work package (WP4)
Start Month:
1
End Month:
48
Lead Beneficiary:
UBAH
Objectives:

Design an established prototype device (EEG/fNIRS/Robotic Stimulator for Study A) and finalised experimentation set- up for Study B. Identify digital biomarkers correlating with sensori-motor domain by demonstrating the efficacy of the multimodal data combination. Conduct robust evaluation of identified stroke biomarker fingerprints through retrospective analysis and establish effectiveness in the lab.

‍

Description:

This Work Package involves designing an advanced multimodal data acquisition pipeline, developing biosignal processing and deep learning models, creating high-fidelity synthetic data enclave and validating multimodal biomarkers (response, predictive, prognostic, monitoring etc). The tasks collectively advance robust, standardised and innovative stroke biomarker discovery and integration.

Task 4.1:

Design considerations for data collection pipeline and hardware

Task 4.2:

EEG/fNIRS Device biosignal processing models, Hyperparameter considerations and optimisation

Task 4.3:

Investigate the impact of deep learning approaches on radiological data

Task 4.4:

Develop and validate generative AI for multimodal data synthesis