Speech Technology Center is implementing a project for filling out medical documents by voice in Moscow

  • Technologies that transform the doctor's voice into the text (speech-to-text) allow healthcare workers to get rid of routine, save 22% of time on filling out protocols, and improve overall documents and diagnosis quality.
  • The solution is already working in the departments of radiological diagnostics of Moscow city medical centers.
  • A successful case scaling up to other regions of Russia.

Speech Technology Center (part of the Sberbank ecosystem) improving the processes of medical institutions, the work of radiation diagnostics doctors using intelligent speech technologies.

At the heart of the solution is Voice2med, an artificial intelligence product – proprietary developed technology by the experts of Speech Technology Center. Voice2med allows simple dictating information for medical cards, protocols and other documents through a microphone. The system recognizes all medical terms and expressions, overall speech recognition quality is 97-98%. During real-life scenario doctor dictates the text, which is recognized in real time and automatically transferred to the final protocol in the medical information system. The solution allows to free hands and save healthcare workers from routine filling forms process and focus on diagnostics.

The pilot project for the implementation of voice input of protocols was carried out in close cooperation with the Moscow City Information Technology Department, after the infrastructure audit the optimal deployment of the system was determined. The Moscow Department of Healthcare has carried out serious methodological work, which made it possible to provide a full cycle of training for medical personnel in specialized training classes. Based on the results of a large-scale pilot project, Speech Technology Center took into account the existing knowledge and improved the final solution to rise the quality of speech recognition to the the highest level.

“Such technology can be a foundation of education process for the residents, so that from the very beginning of work they get used to the fact that they can simply dictate the text of the protocol. As a mandatory part of the basic trainings, this technology definitely will be in demandby the doctors. As a result, such approach will begin to scale. I think this will further contribute to what has already begun to happen - the standardization of medical documentation, the release of the doctor time for a more detailed analysis of diagnostic images, the optimization of the doctor's work, and a reduction in the time for preparing final protocols. Specialists will have more time to analyze diagnostic images and medical documents of patients, ” said Sergei Morozov, chief specialist in radiation and instrumental diagnostics of the Moscow Healthcare Department, director of the Diagnostic and Telemedicine Center.

“Today, when medicine is under tremendous stress, modern technologies and solutions based on artificial intelligence come to the rescue, and Voice2med is definitely one of them. The solution allows to save medical workers from routine, save more than 20% of time on filling out protocols, improve their quality, and most importantly, focus on analysis. A feature of the project is our close interaction with the Department of Information Technology, which made possible to qualitatively test and integrate the solution into existing IT infrastructure as a centralized complex solution. In addition, the Moscow Department of Healthcare provided major methodological support in this project, providing high-quality training for medical personnel at a high level. We hope to develop our projects in order to ensure a new level of service quality and rapid technological support for healthcare, one of the key country industries, ” comments Dmitry Dyrmovsky, CEO of Speech Technology Center.

The greatest value in the use of voice input achieved when describing diagnosis over computed tomography, magnetic resonance imaging and digital radiography. This is due to the fact that short templates of fluorography and mammography protocols are filled much faster with a few mouse clicks. The Center for Diagnostics and Telemedicine Technologies notes that the introduction of a voice filling system for medical protocols can become an effective tool not only for radiologists, but also for surgeons, endoscopists and medical forensic doctors, and in the long term also for therapists and neuropathologists.

Speech Technology Center (part of the Sberbank ecosystem) is a global developer of intelligent speech technologies, face recognition, a technology expert in the field of artificial intelligence and machine learning. One of the few companies in the world that creates and develops both modalities: face and voice biometrics. Technologies for detecting voice falsification (liveness) and speech recognition from scores leading positions in the world ratings of NIST, ASVspoof Challenge, VOiCES, CHiME Challenge. Speech Technology Center solutions are in demand in 70 countries around the world.

Official website: www.speechpro.com.

Media contacts: pr@speechpro.com