Сравнимо с работой опытного врача: томские учёные разработали программу для анализа жировой ткани сердца

The EPIFAT program allows processing computed tomography data in 22.3 seconds, highlighting epicardial adipose tissue with high accuracy

Scientists from the Research Institute of Cardiology of the Tomsk National Research Medical Center and Tomsk Polytechnic University have created the EPIFAT program, which automatically analyzes CT images of the heart, determining the volume and density of epicardial adipose tissue. This software simplifies the diagnosis of cardiovascular diseases.

EPIFAT program interface: results of automatic analysis of epicardial adipose tissue

The EPIFAT program allows processing computed tomography data in 22.3 seconds, highlighting epicardial adipose tissue with an accuracy of 0.954, which is comparable to the work of an experienced doctor. The traditional method requires manual analysis of multiple CT images, which takes a significant amount of time and requires high qualifications. EPIFAT automates the process, making it fast and accessible to medical institutions.

Epicardial adipose tissue plays an important role in the heart's function, participating in metabolic processes and protecting the myocardium. Its changes often indicate the development of cardiovascular diseases. The program uses artificial intelligence algorithms in Python to process data, segment the heart, and calculate the parameters of adipose tissue.

Computed tomography of the heart with analysis of epicardial fat using manual tracing of the epicardium on images - from the archive of researchers at the Research Institute of Cardiology of the Tomsk National Research Medical Center

According to Konstantin Zavadovsky, Head of the Department of Radiation Diagnostics at the Research Institute of Cardiology, EPIFAT is suitable for both scientific research and clinical practice, contributing to the early detection of pathologies and correction of therapy. The program is protected by RF patent No. 2025610317 dated January 9, 2025. Details are published in the collection "Prospects for the Development of Fundamental Sciences."

Read more materials on the topic:

Open system for diagnosing lithium-ion batteries created at NSU

Transition to domestic software caused massive failures in the operation of the portal of state acts

New Antimicro program from St. Petersburg State University will speed up the selection of antibiotics for patients