How new AI technology could save cancer patients
Doctors go through extensive training to learn to diagnose critical conditions such as cancer. But even the smartest, most diligent physicians make mistakes sometimes. According to the healthcare journal BMJ Quality and Safety, breast cancer is misdiagnosed up to 28 percent of the time. Some of these errors occur when physicians are reviewing breast biopsy samples. In the best conditions, pathologists can accurately identify breast cancer in these samples with 96 percent accuracy.
96 percent is pretty good, right? Objectively, yes, but we don’t think it’s good enough for the other four percent. That’s why we’re intrigued about a new technological development that could raise that number even high. Read on to learn more.
Paging Dr. Robot…
Researchers from Harvard Medical School (HMS) and Beth Israel Deaconess Medical Center (BIDMC) recently published a study about an artificial intelligence system trained to detect breast cancer. By “showing” the system hundreds of cancerous and noncancerous lymph node biopsies, they were able to teach it how to accurately detect cancer 92 percent of the time.