Neural network from IBM predicts breast cancer a year before its manifestation (3 photos)

The penetration of AI technology into medicinesignificantly affects the early diagnosis of such serious ailments as diabetes, Alzheimer's disease, heart disease and others. Early diagnosis allows you to effectively deal with diseases, increases the chances of complete healing. IBM engineers have announced the development of a program based on AI technology that can diagnose breast cancer in its early stages with a probability of 87% a year before it appears.

AI technology has already been used foroncology prediction and based on the processing of mammogram data or medical records. An application created by IBM developers uses a hybrid approach that processes data and mammograms and doctors' records. This approach improves the reliability of forecasting.

IBM technique is based on teaching AIvarious types of mammograms, their relationships with biological markers and data from medical records. This approach improves prediction accuracy and reduces the likelihood of misdiagnosis. IBM IBM also obtains additional data for analysis based on biopsy information, laboratory tests, a history of cancer and other sources.

AI algorithm will not be the main and onlytool for predicting cancer. This will be an additional information tool that will allow specialists to identify the patient's predisposition to breast cancer on time. The neural network will help to confirm the diagnosis of the radiologist and start treatment without additional examination, which is very important in case of rapid disease or lack of equipment and specialists for additional analyzes.

Source: engadget.com