IBM's IBM Predicts Breast Cancer A Year Before It Appears

The algorithm has been trained on hundreds of thousands of non-personified data on mammography and clinical

patient histories. The algorithm differs from similar neural networks in that it analyzes both data sets, and not one of them.

The resulting artificial intelligenceable to notice as obvious prerequisites for the emergence of breast cancer, as well as small details - for example, a lack of iron or problems in the thyroid gland.

The researchers improved the accuracy of the algorithm by adding data on biopsy, laboratory tests, registries of cancer patients, and more to its knowledge base.

Earlier it was reported that the next health crisis could be caused by a huge amount of spam from robots, which would not allow real patients to reach the hospitals.