AI endoscopic system IDs colorectal polyps

Researchers have developed an endoscopic system, powered by artificial intelligence (AI), capable of identifying colorectal adenomas during a colonoscopy. Findings are set to be presented at the 25th UEG Week in Barcelona.

The study, led by Yuichi Mori from Showa University in Yokohama, Japan, showcased the endoscopic system's ability to detect colorectal polyps with high levels of sensitivity and accuracy. The system is comprised a computer-aided diagnostic system that used a 500-fold magnified endocytoscopic image to identify more than 300 features of a colorectal polyp. The system then uses AI, which has been trained with more than 300,000 endocytoscopic images using machine learning, to predict the lesion's pathology.

“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. This allows the complete resection of adenomatous polyps and prevents unnecessary polypectomy of non-neoplastic polyps,” said Mori. "We believe these results are acceptable for clinical application and our immediate goal is to obtain regulatory approval for the diagnostic system.”

In the study, researchers enrolled 250 patients with colorectal polyps to test the AI system by comparing pathology reports. Of the 306 polyps evaluated, the AI system was able to reach a sensitivity rate of 94 percent, specificity of 79 percent, accuracy of 86 percent, and positive and negative predictive values of 79 percent and 93 percent.

“Precise on-site identification of adenomas during colonoscopy contributes to the complete resection of neoplastic lesions" said Mori. "This is thought to decrease the risk of colorectal cancer and, ultimately, cancer-related death."