[an error occurred while processing the directive]
RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2
2021/12/21 17:36:12

Digital consultation: how to create AI services in medicine

Tags: Artificial Intelligence, Medicine, Healthcare, IT

Contractors: SberMedII

Product: Medical Digital Diagnostic Center - MDDC (Medical Decision Support Platform) -- >

Content

Understand the need for a doctor and offer a digital solution

Building a service using artificial intelligence (AI) for medicine always begins with identifying and understanding a problem that is relevant to the medical community. At this stage, it is important to build communication between developers and experts, whose roles are acting doctors of medical institutions. The latter have their own tasks that require solving by means of modern technologies. Having learned about them from doctors, the developers report whether it is technically possible to create a service and if so, they offer it to doctors. Together, they make the decision to start development.

For example, there is a problem: radiologists have to spend a lot of time examining CT scans of the brain when a stroke is suspected (usually from 15 minutes to 1 hours). But everyone knows how important the time factor is in this case. The faster the diagnosis is made, the greater the chances for the patient to avoid severe consequences in the form of, for example, disability. A new, "breakthrough" solution was required, doctors shared this problem with the developers and as a result, the CT Stroke service was created. And already today, the development can help medical specialists in their daily work. Based on AI algorithms, the model automatically marks CT images, determining the type and location of the stroke zone. The processing time for each image was reduced to several minutes.


The most common example of the use of AI in the past year to medicine was the processing computer of tomography images of the lungs affected. COVID-19 The percentage of lesion affects what treatment a patient should receive. Identifying the share of this defeat is always painstaking and difficult work, even for an experienced doctor. At the same time, AI can be processed quickly and efficiently, literally in a matter of minutes, data thereby helping the doctor in making the final decision and making a diagnosis.The CT of the Lungs service for the analysis of CT images has shown its effectiveness during a pandemic and is successfully used now.

He processes each picture for several minutes, in the most difficult cases - no more than 15 minutes. In addition, the service indicates priorities for doctors: they are primarily provided with images with suspected pathology. That is, you can quickly start treatment in those patients who need it most.

Thus, today's developments in artificial intelligence respond to the current needs of the medical community and are designed to help them in their work.

Data: find and depersonalize

At the next stage, to create a service, you will need to obtain data and expertise from doctors, on the basis of which machine training will begin. This is similar to how older colleagues teach less experienced physicians by sorting out cases from clinical practice. The only difference is that machine learning requires an order of magnitude more data - not even hundreds and thousands, but hundreds of thousands of patients. And these should not be just impersonal patient data, but specific cases analyzed and confirmed by the doctor.


But first you need the data to be correctly collected and marked up. On untrained data, it is impossible to develop an effective AI model. Therefore, it is important to choose specialized clinics with which negotiations are being held, for what tasks anonymized data will be collected. Next, computational experiments start, which have the goal of choosing the optimal architecture and setting parameters for a specific task. Collecting data, checking quality at each stage and marking it in any medical institution is a very laborious process, it requires the involvement of the most qualified doctors. And technicians should bring all collected data to a "single denominator," eliminating distortion. It is important that the collected arrays are homogeneous in most parameters, and the samples are balanced (for example, that the number of cases with pathology and without pathology be approximately the same).

After all, if, for example, the model is trained in pictures taken on one equipment, and then it meets with images taken on another equipment, this can lead to an error. Simply because the model will see some new features of the data and may erroneously interpret them. There are a lot of manufacturers and modifications of equipment, which is why even those AI models that are already used in industrial operation are constantly updated, completed - this is a constant, incessant process. That is, they improve their qualifications, as is customary in the medical community. One example of working with datacet is the creation, together with the Moscow Government, of a smart assistant to the TOP-3 doctor.

In Moscow, a large study was conducted based on data from anonymized visits to medical institutions. Data analysts summarized the statistics of all diagnoses in outpatient treatment for one year in the Moscow region, analyzed more than 2.2 million visits of 420 thousand patients. After that, they ranked the probability of the disease category according to the generally accepted classification MKB-10 and selected 265 categories that cover 95% of the sample cases. As validation data, 1.7 million visits of 700 thousand patients not previously used were taken. As a result, the Smart Doctor Assistant service was created, which became the champion of the prestigious WSIS Prizes 2021 international competition, held under the auspices of the UN. At the entrance, the "assistant" receives the text from the medical record, the history of primary hospitalization, then a mathematical model based on neural networks works, and it issues the three most likely diagnoses according to the International Classification of Diseases (ICD-10). Now it is used in all adult polyclinics in Moscow, as well as in a number of regions of Russia.

There are many services - one platform

In order for artificial intelligence to quickly enter the arsenal of a modern doctor, it is necessary that it fits as organically as possible into existing working interfaces. Conditional comparison: if a person uses a calculator, it will be very strange if he is offered another separate device ("with artificial intelligence"). But he is unlikely to be surprised if the new calculator models solve increasingly complex problems. The task of medical AI manufacturers is to reduce the burden on the healthcare system in general and on certain areas of medicine in particular. However, this, of course, will not be possible immediately (as doctors did not immediately get used to using ultrasound scanners and CT machines, which were also once "new"). That is why new software with AI is being introduced into information systems familiar to doctors, complementing their functionality. As a result, doctors do not need to specifically learn, breaking away from everyday work in order to start working with artificial intelligence, enough common basics of computer literacy. And most often, medical institutions need not some separate service, but a set of algorithms that can be used in a particular situation.

So came the idea of ​ ​ creating a Medical Digital Diagnostic Center (MDDC). It is a comprehensive solution to help make a diagnosis using artificial intelligence. However, in any case, it is validated by specialist doctors. MDDC combines more than 50 products and solutions of SberMedII, as well as other companies of the Sber ecosystem and partners.

Doctors of clinics connected to MDDC can send primary admission, instrumental and laboratory diagnostics to the "digital consultation" online. Digital documents can be downloaded through a web interface or a separate application for an automated workplace, or through programs already integrated into the clinic's medical information system (MIS).

Information is automatically routed and processed by AI algorithms, preliminary conclusions are submitted for verification to relevant MDDC doctors, and after that the recommendation for diagnosis is returned to the clinic. The final recommendations are issued by MDDC doctors specializing in a wide profile of medical areas: therapy, cardiology, radiology, oncology, surgery, dental research.

Thus, modern technologies can effectively and efficiently help doctors in their daily work. AI algorithms help significantly speed up the processes of making medical decisions, do not miss an error, pay additional attention primarily to the areas of possible pathology (prioritization of the list of patients). An extremely important area is the possibility of retrospective analysis of medical data using artificial intelligence.