Face recognition to support pre-hospital stroke diagnosis.

Stroke represents one of the most important killer after heart disease and cancer and it is the leading cause of severe long-term disability with huge health and social costs. About 85% of strokes are ischaemic and in these cases the faster blood flow is restored to the brain the better chance someone has of making a good recovery. Early detection and treatment of stroke is critical to saving lives and reducing permanent damage.In this paper, the first results about the adoption of Convolutional Neural Networks to identify the symptoms of the stroke and support the pre-hospital medical diagnosis are illustrated. Starting from the pre-hospital diagnosis of the stroke using the FAST test or other scales (e.g. Cincinnati Prehospital Stroke Severity Scale or National Health Institute Stroke Scale), based on the analysis of well-defined signs on the patient’s facial expressions and movements, the study is currently focused on facial expressions.