ST. PETERSBURG, March 5 (Xinhua) -- Russian researchers have developed a neural network that can detect early-stage breast cancer from CT scans within minutes, significantly faster than traditional methods that may take up to a day, local media reported on Thursday.

The technology, created by St. Petersburg Electrotechnical University and the Almazov National Medical Research Center, reduces clinical error probability by about 20 percent, according to developers.

"Such a development could help save thousands of lives," said Denis Stepanenko, head of the OncoDetect AI project and a master's student at the university's Faculty of Computer Technologies and Informatics.

Breast cancer claims 4,000 to 7,000 lives annually in Russia, with late diagnosis cited as a major contributing factor. The AI system automatically processes CT images and highlights potential cancer signs for further evaluation by physicians.

Developers said the model was trained on around 7,000 anonymized CT images from the Almazov Center's database. ■