IBM’s neural network predicts breast cancer a year before its onset (3 photos)

The penetration of AI technology in medicine significantly affects the early diagnosis of such serious ailments as diabetes, Alzheimer’s disease, heart disease and others. Early diagnosis allows you to effectively fight 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 the early stages with a probability of 87% a year before its appearance. AI technology has already been used to predict oncology and was based on the processing of mammograms or medical records. The application, created by developers from IBM, uses a hybrid approach, which processes data from mammograms and doctor’s records. This approach increases the reliability of forecasting. The IBM methodology is based on teaching AI various types of mammograms, their relationships with biological markers and data from medical records. This approach increases the accuracy of prediction and reduces the likelihood of an erroneous diagnosis. IBM AI also receives additional analysis data based on biopsy information, laboratory tests, the historical cancer registry, and other sources. AI algorithm will not be the main and only tool for predicting cancer. This will be an additional information tool that will allow specialists to timely identify the patient’s predisposition to breast cancer. The neural network will help confirm the diagnosis of the radiologist and begin treatment without additional examination, which is very important in case of a fast-moving disease or in the absence of equipment and specialists for additional analyzes.

Sam Richards

Add comment

Follow us

Don't be shy, get in touch. We love meeting interesting people and making new friends.