The project derives its name from the digital microscopic analysis laboratory at Moscow State Medical University. Sechenov collaborated with researchers from Yaroslav the Wise Novgorod State University to teach artificial intelligence how to recognize blood and lymphatic vessels that contain tumor cells in histological sections of lung cancer patients. This work was discussed with socialbites.ca at Sechenov University.
The effort aims to reduce misinterpretations when assessing tumor spread and to support the selection of appropriate treatment strategies, ultimately improving patient outcomes.
Lung cancer remains a leading cause of cancer mortality. Each year, clinicians identify more than two million new cases. A key unfavorable factor is vascular invasion, a process where tumor cells extend beyond the primary tumor, travel along vessel walls, and disseminate through the body.
According to Anna Timakova, a junior researcher in the Laboratory for Digital Microscopic Analysis, the presence of vascular invasion raises the risk of distant metastasis and worsens overall survival. If such regions appear in histological slides, additional treatments such as radiation therapy may be considered after surgery. Detecting these areas, however, is challenging.
Researchers trained a model on scans from nearly 200 micro slides to identify vessels in histology images. The method currently achieves a sensitivity of about 75 to 80 percent.
The development could enable the detection of vessels that harbor tumor cells within their walls but have not yet breached the lumen in the examined section. The medical community has not reached a consensus on whether these suspicious regions should influence future treatment plans. Nevertheless, automating their detection with machine learning offers clinicians a chance to reassess the true prognostic value of such controversial areas.
By year-end, the team plans to share performance metrics on specificity and sensitivity as part of an internal dataset. In the following year, the model will be evaluated on external datasets and real patient cases.
Previous Russian research highlighted a crucial anti cancer mechanism in uterine tissue, underscoring the broader role of vascular biology in cancer care. At Sechenov University, ongoing investigations continue to push the boundaries of how AI aids histology and oncology.