[ad_1]
Copy link
Scientists at the Polytechnic University of St. Petersburg have developed a program that can accurately diagnose the presence of cancerous tumors in the lungs in less than 20 seconds, by analyzing the data extracted from the patient's lung scanner, taking count other analyzes.
The system was named Dr. Azimov in honor of science fiction author Isaac Azimov, who invented three laws of robotics.
According to the university website, the program can be placed on any computer, where the system analyzes the data downloaded from the patient's chest tomography, taking into account other analyzes. . Once the information is processed, the program publishes a report on the diagnosis of the disease.
"In the beginning, we developed an algorithm to look for lesions of 6 mm, because the doctors themselves adhere to this tactic, but this system is so well studied that it detects new tumors of smaller size", said Lev Utkin, project director.
The system is also capable of self-learning, a feature based on a neural network capable of taking into account new data, and each resolved task speeds up the process in the future. It is assumed that after "training", the program will be able to achieve the result in two seconds.
[ad_2]
Source link
