Hybrid intelligence system for the evaluation of upper extremity movement quality in stroke survivors.
Installation at the highest ranked rehabilitation hospital.
Used for data capture of 100 participants at the Shirley Ryan Ability Lab, Chicago, Illinois.
Graphic user interface layout for data data capture.
ARAT kit, optimized for simultaneous human observation and computer vision.
Camera layout allowing for spacial reconstruction and movement quality assessment.
Full system setup, accessible by wheelchair bound participants.
Graphic portraying the transformation of tacit clinical knowledge to explicit information as the system evolves through human and computational input.
Illustrated significance of applying a standardized approach in conjunction with noisy datasets to achieve computation in complex scenarios.
Breakdown of complex tasks through segmented movement features.
Demonstration of the variability of assessment reasoning at the level of the impaired feature.
Development of weighted feature significance based on clinical observation.