AI software can now detect bowel cancer in under a second

Bowel cancer can now be detected within seconds using artificial intelligence software recently developed in Japan, researchers have claimed.

A team of researchers from Showa University in Yokohama has developed a computer-aided diagnostic system that can detect cancerous polyps in endoscopy images in less than a second.

The system uses a 500-fold magnified view of a colorectal polyp to analyse nearly 300 features of the polyp after applying narrow-band imaging (NBI) or staining with methylene blue.

Three hundred and six images of colorectal polyps, taken from the bowels of 250 women and men, were presented to the AI system. The samples presented contained both cancerous and non-cancerous polyps.

The system was used to predict the pathology of each polyp. These predictions were then compared with the reports obtained from the



final resected specimens.

The system was able to correctly identify 94 percent of the cancerous samples, taking less than one second for image analysis.

According to Yuichi Mori from Showa University , the most remarkable breakthrough with this system is that artificial intelligence enables real-time optical biopsy of colorectal polyps during colonoscopy, regardless of the endoscopists' skill.

The results of the study have been presented at United European Gastroenterology Week in Barcelona, Spain.

 Mori said,we believe these results are acceptable for clinical application and our immediate goal is to obtain regulatory approval for the diagnostic system.

The team is presently working on a multi-centre study to achieve this and is trying to develop an automatic polyp- detection system.