The latest artificial intelligence system is expected to improve the efficiency of colorectal cancer screening：Although colorectal cancer is the third most common cancer in the United States, it can be prevented to some extent. If doctors can find those precancerous lesions early, they can remove them before the problem occurs. Colonoscopy is very effective in detecting precancerous polyps and preventing colon cancer, but the effectiveness of this procedure depends largely on the ability of the executive physician.
In his 20 years as a gastroenterologist, Michael Wallace has performed hundreds of colonoscopies. He is good at identifying inflammation or polyps that may evolve into cancer along the colon bulge, but some flat polyps are difficult for doctors to see with the naked eye.
Now, the U.S. Food and drug administration has approved a new tool, the artificial intelligence system GI genius, which is expected to help doctors identify the growth of precancerous polyps during colonoscopy. It can see more inside the colon than most doctors.
During the drug test of the system, more than 13 million colonoscopy videos were carried out in Europe and the United States, and the algorithm was trained to recognize polyps. In order to train AI to distinguish potential diseases, gastroenterologists mark these images as normal or unhealthy tissues, and then conduct a step-by-step difficult to identify polyp test on AI, starting with colonoscopy under ideal conditions, and then moving to a more difficult challenge such as distinguishing small polyps, similar to a hidden black spot. The system can be added to the oscilloscope used by doctors for colonoscopy. When detecting the colon, a green box is used to mark the potential polyps.
GI genius was approved in Europe in October 2019. It is the first artificial intelligence system approved by the U.S. Food and drug administration to help detect colonic polyps.
A gastroenterologist at New York University said that it makes sense that AI is good at identifying polyps, because millions of colonoscopy videos provide a large amount of data to make the algorithm comprehensive, which can protect the system from the bias in other health care algorithms. “This technology should help us do better, but it can not replace the highest quality technology we can provide. I don’t think this is a revolution, but this is the beginning of artificial intelligence to help us,” he said