You cannot postpone the implementation of AI in medicine / News of the Society of Krasnoyarsk and the Krasnoyarsk Territory / Newslab.ru

“The massive integration of AI services is able to significantly reduce the time that doctors spend on routine tasks,” said Sergey Karpovich, deputy head of the T1 AI (IT Holding T1), at the “Robotized Treatment and II Nagation systems” at the Kazan Digital Week forum.

According to him, the active implementation of modern and technologies in medical institutions is the key to the rapid economic efficiency of digitalization and improving the quality of medical care. “

Today are health workers spend up to 50 % of the time for routine tasks: compilation of reports, maintaining statistical accounting, patient recording and other administrative processes. At the same time, according to the doctors themselves, up to 80 % of the working day can take the documentation. This reduces the effectiveness of patients and can affect the quality of diagnosis and treatment. The key trends in the use of AI in medicine are already changing the situation. Thus, healthcare organizations use speech analysts, introduce voice assistants and generative AI for processing documents, as well as integrate neural networks into business processes of medical institutions.

The use of speech analytics and voice assistants in Russian medical institutions allows doctors to save doctors up to 2 hours a day, thereby increasing the time to work with patients by reducing the volume of manual work at the computer. In particular, Sergey Karpovich noted that modern models that can recognize and differentiate voices reduce the time to record and fill out the inspection protocol up to 30 seconds, while earlier this process took about 5-7 minutes. And the use of RAG services-solutions that combine language models with the base of medical knowledge-allows you to generate formulations corresponding to clinical recommendations. At the same time, such and solutions can easily integrate into the existing interface of the medical information system (MIS).

According to the speaker, the issue of automation of medical documentation and business processes, in particular in organizations operating under the compulsory medical insurance system, is also especially acute. Errors in the documents often lead to no more than 30 % of insurance payments. The introduction of AISSisters will automate the verification of medical documentation for compliance with regulatory acts and instructions, analyst financial efficiency, forecasting planned costs and key indicators of activity.

“Artificial intelligence is an important stage in the development of medical technologies: neural networks help to improve the quality of medical services, today the integration of the simplest AI solutions shows high results. If we resist the mass implementation of and technologies or postpone the study of effective approaches to their use, in the future this will lead to digital technological lag and additional financial investments. Therefore, it is important to move at the pace of development of AI, consider it as a reliable assistant and use this tool, ”said Sergey Karpovich.

Nevertheless, the expert emphasized that the introduction of artificial intelligence is a complex and long process, which on average takes about six months from the date of setting the task until the finished II service. Moreover, it may not be advisable to implement such a project within one medical institution. That is why ready-made platform IT products and cloud services are an effective solution-they can be integrated with the existing turnkey medical system in just a few weeks. At the same time, such complex vertical and solutions make it possible to automate processes not only quickly, but also without significant financial costs.