Pilot implementation of speech recognition technology in the endoscopic centers of Moscow Healthcare Department

Abstract

The endoscopic centers (EC) are faced with intensive overload due to the high patient flows, and the solution might be to implement into routine practice the speech-to-text software for inputting endoscopic examination descriptions. The authors assessed the changes in the time of paper work when endoscopists used speech recognition software, modern wireless headset, available Health Information System, algorithms and standardized protocols. Endoscopists were able to fill in the examination protocols using the developed method of speech recognition based on artificial intelligence (AI) and Voice2Med software. AI-based assessment shows that the absolute majority of professional terminology from the endoscopists' speech was reproduced (the recognition error was less than 3%). Moreover, 100% of endoscopists were satisfied with the voice input technology for making protocols. The study demonstrated that the protocols of endoscopic examinations developed as the result of the project allow endoscopists to reduce the time for writing medical reports as well as the overall time of procedure. For citation: Shabunin AV, Bedin VV, Korzheva IYu, Orlov SYu, Osmak EK, Vasiliev IV, Nesterov DV, Alepko AA. Pilot implementation of speech recognition technology in the endoscopic centers of Moscow Healthcare Department. City Healthсare. 2023;4(1): 68-74 doi: 10.47619/2713-2617.zm.2023.v.4i1;68-74.