Impact of Viz.ai and RapidAI on Thrombectomy Times in Large Telemedicine System
Theresa Sevilis1, Mariecken Fowler1, Amanda Avila1, Caitlyn Boyd1, Lan Gao2, Oleg Collins2, Thomas Devlin3
1TeleSpecialists, LLC, 2University of Tennessee- Chattanooga, 3University of Tennessee Health Science Center
Objective:

Compare the arrival to Neurointerventionalist (NIR) Notification times for facilities utilizing Viz.ai, RapidAI, and no AI software for acute stroke.

Background:

The utilization of artificial intelligence (AI) platforms to provide rapid access to images, LVO detection, and team communication in acute stroke protocols is expanding globally. There is evidence from small studies that they enhance care, but minimal evidence from large trials directly comparing stroke workflow metrics of these platforms.

Design/Methods:

Acute stroke consultations seen in the emergency department in 227 facilities (27 states) from July 1, 2021 to December 31, 2021 were extracted from the TeleCare by TeleSpecialistsTM database. The encounters were reviewed for demographics, LVO, accepted for intervention, arrival to NIR notification time, and AI software used. Patients were classified into three groups based on use of Viz.ai, RapidAI or no AI software. The median arrival to NIR notification times were compared.

Results:

A total of 25,430 patients were included: No AI group (7,433), RapidAI (6,726), and Viz.ai (11,271). The median arrival to NIR notification time for No AI was 89 minutes when compared to RAPIDAI 58 minutes and Viz.ai 54 minutes was significantly longer, p <0.0001. There was no significant difference when RapidAI and Viz.ai were directly compared. More advanced imaging was performed and LVOs detected with AI software platforms when compared to No AI. RapidAI had a higher rate of advanced imaging performed and more LVOs found than Viz.ai.

Conclusions:

Utilization of an AI platform showed earlier notification of NIR by more than 30 minutes in this large multi-state telestroke system. No significant difference was seen between platforms for time to NIR notification. Since the utilization of either AI platform had such a profound impact, this is strong evidence to recommend utilization of an AI platform at all stroke centers.

10.1212/WNL.0000000000205476