A system of voice filling of ultrasound protocols is being introduced in Moscow clinics

2021: Introduction of voice filling systems for ultrasound protocols in hospitals in Moscow

In mid-April 2021, it became known about the appearance in Moscow polyclinics and hospitals of a system of voice filling of ultrasonic examination protocols (ultrasound). The technology allows you to dictate the description of the study and not waste time entering information manually.

According to Anastasia Rakova, deputy mayor of the capital for social development, it takes only a few days for doctors to master the new tool, and its constant use allows you to significantly reduce paperwork time and pay more attention to research: tests show that doctors using voice input spend an average of 20% less time to complete the protocol.

A system of voice filling of ultrasound protocols is being introduced in Moscow clinics

The voice assistant was created on the basis of artificial intelligence and can constantly be trained: to replenish the dictionary with complex and rare medical terms, to adapt to the pronunciation of different doctors, to cover new diagnostic and therapeutic tasks.

Rakova in April 2021 noted that voice technology was previously tested by ultrasound doctors of six clinics, and in the near future it will be available in other medical institutions.

The digital assistant automatically decrypts the spoken words, translates them into text format and inserts them into the protocol template. After the radiologists began to successfully use it, we decided to start using the service in the work of other specialists, "she said.

As noted on the official website of the Mayor of Moscow, the introduction of such technologies was made possible thanks to a single digital healthcare platform, which was jointly developed by the Moscow Social Development Complex and the Department of Information Technology. Artificial intelligence-based services are also used in radiology - they are integrated into the Unified Radiological Information Service.^[1]

2020: Radiologists in Moscow began to draw up conclusions in the speech recognition system

On June 10, 2020, it became known that radiologists in Moscow clinics began to make conclusions in the speech recognition system. About the introduction of voice input technology  at the conference "Radiation Diagnostics of Moscow: 2019-2020" said the chief specialist in radiation and instrumental diagnostics of the Moscow Department of Health (DZM) Sergey Morozov.

According to him, voice input is most effective in describing studies of computer and magnetic resonance imaging and digital radiography. This is due to the fact that short templates of fluorography and mammography protocols are more quickly filled with several mouse clicks.

Radiologists in Moscow clinics make a conclusion in the speech recognition system

The voice input system was developed jointly with the Center for Speech Technologies (MDG). To improve the system, the Center for Diagnostics and Telemedicine of the Moscow Department of Health transferred 2.6 million anonymized X-ray protocols to the MDGs.

Sergey Morozov cited data from timing studies, according to which, using voice input, the radiologist spends 22% less time on preparing the protocol. The Center for Diagnostics and Telemedicine states that by June 10, 2020, the system is trained to recognize speech with abbreviations, abbreviations and specific medical terms.

As Morozov noted, during the pilot tests of the development, the accuracy of speech recognition was at the level of 87%, and by June 10 the figure reached 93%. The main errors are misrecognized prepositions and inconsistent sentences. After completing the protocol, the doctor re-reads and corrects the incorrectly recognized words, he said.

The cost of connecting one microphone to the workplace of a radiologist is estimated at 20 thousand rubles. The total costs of the project are not disclosed.^[2]

Notes