Machine learning system accurately predicts survival across 10 different cancers

A Google Health (CA, USA) funded machine learning system has predicted disease-specific survival across ten different types of cancer with a high degree of accuracy. The technology uses histopathological images to stratify disease risk on an individual basis.Predicting survival and prognosis in cancer has important implications for treatment decisions and monitoring, and can therefore have an effect on disease progression. The research team aimed to improve the accuracy of this information via the machine learning system, so clinicians can provide the most appropriate treatments for patients. Cancer prognosis and survival predictions are currently based on a wealth of baseline information,...