FIELD: computer technologies.
SUBSTANCE: present solution relates to medical decision support systems using artificial intelligence technologies. Disclosed is an intelligent medical decision support system, wherein the system provides: obtaining input data from an electronic medical record (EMR) of a patient, which includes data selected from the group: at least one examination protocol presented in text format or in tabular format, parametric data, parametric data array, at least one medical image or combinations thereof; processing input data using a preprocessing module, which comprises: a text data processing module based on a machine learning model, providing data processing in natural language and determination of symptoms of at least one disease; medical image processing module, comprising: a data classification model, which classifies incoming images and determines the type of at least one machine learning model processing the corresponding anatomical region selected from the cascade of models; cascade of machine learning models, each of which performs detection and extraction of symptoms of at least one disease based on processing of corresponding anatomical region on incoming medical images; generating a list of symptoms of at least one disease of the patient, obtained based on the processing of the data by the preprocessing module; processing of the obtained list of symptoms using a module for determining nosologies, which provides diagnosis and prediction of diseases of the patient, wherein the nosology determination module comprises a concurrently operating nosology search module based on an algorithmic method, and a nosology search module based on a machine learning model.
EFFECT: invention provides higher accuracy of automated diagnosis, including clarification of the diagnosis of the underlying disease, its complications and comorbidities, as well as the possibility of predicting the risks of developing other pathologies.
6 cl, 7 dwg, 2 tbl
