Abstract

Artificial Intelligence-Directed Emergency Stroke Treatment – The VitalFlow Faucet

Author: Mark Borsody

Our research consortium has developed a non-invasive, easy-to-use medical device that can reverse the loss of blood flow to the brain causing ischemic stroke. The device, called the VitalFlow stimulator, takes 15 minutes to administer and can be used immediately in the ambulance if only ischemic stroke patients could be distinguished from patients with hemorrhagic stroke and stroke-mimic conditions (e.g., seizures). Since there are nowhere near enough physicians to cover the 148,000 ambulances serving just the U.S. and Europe, we must build the diagnostic capability into the VitalFlow device itself using artificial intelligence (AI). We now intend to develop an AI diagnostic algorithm enabled with natural language processing that can evaluate patients in the ambulance setting and that can accurately determine the diagnosis of ischemic stroke with limited need to engage on-call physicians via a video conferencing link. We call the AI diagnostic algorithm “the Faucet”. We plan that the Faucet will be based on IBM’s Watson™, although other AI platforms might be comparably suitable. The Faucet will initially learn to diagnose ischemic stroke based on (1) the real-time evaluations of stroke neurologists provided via a video conferencing link to the VitalFlow devices and (2) existing databases of stroke patient characteristics, such as the > 1 M patient records of the American Heart Association’s Get With The Guidelines Stroke database. Over time, as the Faucet’s machine learning abilities improve its diagnostic accuracy, it will graduate to providing the majority of VitalFlow treatment decisions in the ambulance under the oversight of stroke neurologists, eventually becoming the first-line diagnostician for routine stroke cases. In this manner, the VitalFlow will be a usable emergency treatment for ischemic stroke across the multitude of ambulances worldwide and it will extend emergency stroke healthcare services into the pre-hospital setting.

Co Author/Co-Investigator Names/Professional Title: Mark Borsody, M.D., Ph.D. Chief Medical Officer, Nervive Inc. Senior Scientist, NeuroSpring Margaret E. Groh, Ph.D., M.B.A. Chief Executive Officer, Nervive Inc. Emilio Sacristan, Ph.D. Chief Medical Officer, Nervive Inc. Professor of Biomedical Engineering, Metropolitan University of Mexico City