Day: March 19, 2018

Computers can spot cancer faster than specialist doctors

Computers can spot cancer faster than specialist doctors

Computers can diagnose prostate cancer almost as accurately as specialists, according to research that promises to give results within hours.

Artificial intelligence software was trained to grade the severity of tumours accurately in an advance that will ultimately mean much quicker diagnosis.

The machine learning algorithm is already more than 99 per cent accurate at assessing biopsy samples and within a few years could be routinely used in NHS clinics. The use of AI to interpret scans and samples is one of the fastest-progressing areas of medicine, and last year Jeremy Hunt, the health secretary, said that within a decade we could be going to computers instead of doctors for a diagnosis. NHS bosses are looking closely at how to use such programmes to interpret x-rays and MRI scans.

 

The first such study suggests that the method will also be a viable way of assessing biopsy samples taken from men thought to have prostate cancer. Chinese scientists took tissue from 283 patients and then divided them into 40,000 samples, 30,000 of which were used to train software to spot those that contained cancer.

The program was then tested on the remaining 10,000 samples and was able to distinguish cancer in 99.38 cases out of 100 when assessed against pathologists. It was also able to classify the tumours by grade, Hongqian Guo, of Nanjing University, told the European Association of Urology congress in Copenhagen last week.

“Until now, automated systems have had limited clinical value, but we believe this is the first automated work to offer an accurate reporting and diagnosis of prostate cancer. In the short term this can offer a faster throughput, plus a greater consistency in cancer diagnosis,” Dr Guo said. “We still need an experienced pathologist to take responsibility for the final diagnosis.”

Jo Martin, president of the Royal College of Pathologists, said that AI would help to ease a shortage of specialists and help by doing routine analysis, freeing up pathologists for unusual cases.