Output details

A Learning Model for the Automated Assessment of Hand-Drawn Images for Visuo-Spatial Neglect Rehabilitation

Establishes a new technique to significantly improve clinical assessment of patients post-stroke. Current assessment to determine the presence and extent of visuo-spatial neglect is difficult in many cases and usually time-consuming for patients (who consequently suffer serious fatigue) and clinical staff. This novel approach, based on intelligent automated analysis of figure-copying tasks, models the performance of the “gold-standard” testing procedure with a significantly reduced test specification. This has underpinned the introduction of new patient evaluation approaches, initially in collaboration with clinicians at the Kent and Canterbury Hospital, but with on-going opportunities for more widespread uptake (contact Dr.J.Potter, Consultant, jonathanmartinpotter@hotmail.com).