AI Platform for Training Pathologists Being Developed at Sechenov University

Digital program with artificial intelligence will help doctors diagnose diseases more accurately

Sechenov University is creating an innovative educational program that is changing the approach to training pathologists. Together with the company "Arkhitektonika," the university is developing a digital platform that uses artificial intelligence (AI) to train students and doctors to work with histological scans. The program, which is scheduled to launch at the end of 2025, will allow for the diagnosis of diseases with high accuracy and train neural networks to recognize pathologies.

Presentation of the development of a unique digital program for training pathologists

The system is based on digital histoscans — images of tissues uploaded from the telemedicine platforms of the university's partners. The database includes anonymized data from thousands of patients at the Clinical Center for Health Sciences. Students and doctors learn to analyze these images, using AI to identify morphological changes. A unique feature of the program is the "Socratic dialogue" with AI, which asks questions about clinical cases and helps students arrive at the correct diagnosis.

In addition, the program includes the so-called Socratic dialogue with artificial intelligence. It asks questions about cases, and, depending on the answers, continues the dialogue, leading the student to the correct answer.
Tatyana Demura, Director of the Institute of Clinical Morphology and Digital Pathology

According to Tatyana Demura, the profession of pathologist is undergoing a digital revolution. Traditional methods of work are supplemented by AI tools that speed up diagnosis and reduce the likelihood of errors.

As a result of using such a unique platform, we are producing specialists of a new type. They will not only be able to immediately work in modern healthcare system programs used in the best pathology departments, but also have clinical thinking. At the same time, this does not cancel face-to-face classes for future doctors – it is impossible to train a pathologist without them.
Tatyana Demura, Director of the Institute of Clinical Morphology and Digital Pathology

The program covers a wide range of medical fields: dentistry, pediatrics, general medicine, and biotechnology. Doctors will learn to mark histoscans for training neural networks, which will become part of the final testing. The coincidence of the markup with the reference markup performed by the teacher will confirm successful completion of the course. The system is already being successfully used for students, and will soon be adapted for additional education of practicing doctors.

Read more materials on the topic:

Method for synthesizing promising europium compounds for sensors and protection against counterfeiting developed by Russian scientists

Aerosol printer for printing microcircuits and neuroimplants developed by Russian scientists

New type of reactor for producing synthesis gas created by scientists from Perm Polytechnic University