AI cancer detectors | Tech News
More accurate than dermatologists
An AI system developed by a team from Germany, France and the US can diagnose skin cancer more accurately than dermatologists. In the study, the software was able to accurately detect cancer in 95% of images of cancerous moles and benign spots, whereas a team of 58 dermatologists was accurate 87% of the time.
An algorithm for prostate cancer
Chinese researchers have developed an algorithm that can diagnose prostate cancer as accurately as a pathologist. Research leader Hongqian Guo of Nanjing University said: “[This] will help pathologists make better, faster diagnoses, as well as eliminating the day-to-day variation in judgment that can creep into human evaluations.”
Targeting head and neck tumours
Researchers at the University of Texas, Houston, have developed software to accurately contour the shape of head and neck cancer tumours. In studies, trained physicians have been shown to make widely varying evaluations of tumour volume. This software enables radiation oncologists to target their treatment more accurately, particularly critical in these types of cancers because of other vulnerable tissue located nearby.
Scientists from Yokohama, Japan, have developed software that can detect colorectal cancer in its earliest stages with 86% accuracy. This type of cancer is often difficult to locate before the tumours become malignant and deadly, as cancerous cells enter the bloodstream, so early detection is critical.
Researchers at Imperial College London are working with DeepMind Health to develop AI-based techniques to improve the accuracy of breast cancer screening. By studying 7,500 anonymised mammograms, they hope to create software that will reduce overdiagnosis and false positives. They aim to report their findings later this year.