Sechenov University and Russian developer RTK Radiology (a subsidiary of Rostelecom PJSC, part of the DigitMed group of companies) have completed the main development phase of a software module for preoperative planning of thoracic surgeries. The software has been tested at Sechenov University clinics.

The module analyzes computed tomography results and automatically performs segmentation of the lungs, bronchial, arterial and venous trees, and identifies and localizes neoplasms. Using interactive analysis tools, surgeons can assess the functional consequences of potential removal of specific lung structures for each patient. According to test results, the quality of lung tissue segmentation reaches a DICE score (segmentation accuracy coefficient) of 93%, bronchial tree branching coverage depth — up to 96%, and vascular tree coverage — up to 97%. The analysis results are saved as a report that can be added to the patient's electronic medical record.

"We view this AI module as a tool for personalized oncological surgery. It enables a shift from visual assessment to quantitative analysis of an individual patient's anatomy and prediction of the functional consequences of surgery. This enhances the validity of surgical decisions and establishes a new standard for preoperative planning," emphasized Professor Marina Sekacheva, Director of the Institute of Personalized Oncology.

"For a surgeon, it is critically important to have a complete anatomical picture before the operation — not only the tumor location but also the topography of blood vessels and bronchi. This approach minimizes the extent of resection, reduces the risk of intraoperative and postoperative complications, and increases the predictability of outcomes," noted Evgeny Tarabrin, Director of the Clinic of Hospital Surgery at Sechenov University.

"The main goal of the module is to reduce risks for the patient by providing the surgeon with an objective, quantitatively substantiated assessment of the specific patient's anatomical features. This is more than just visualization — it is a prediction of postoperative respiratory function and a calculation of functional losses after resection. This approach is especially important in complex individual cases where standard protocols prove insufficient," added Alexey Bychenkov, Deputy Technical Director of RTK Radiology.

The module was developed in collaboration with practicing thoracic surgeons and radiologists. The system operates on the basis of the DIGIPAX Unified Radiology Information System and can be used in medical institutions in the regions where it is implemented. An active process of submitting the module for registration in the Roszdravnadzor (Russian Federal Service for Surveillance in Healthcare) medical device registry is currently underway. After obtaining the registration certificate, the solution is expected to be available for implementation in federal and regional clinics.