This AI can help eliminate colon cancer


There’s Michael Wallace During his 20 years as a gastroenterologist he performed hundreds of colonoscopies. He thinks he’s pretty good at recognizing growths or polyps, it can flow through the pores of the colon and turn into potential cancer. But he is not always perfect. Sometimes the polyps are flat and hard to see. Other times, doctors just miss them. “We’re all human,” said Wallace, who works at the Mayo Clinic. “We’re tired,” he said, referring to the minute details for the one-minute back-to-back processes.

Colonoscopies, if unpleasant, are Highly effective To detect pre-cancerous polyps and prevent colon cancer. However the effectiveness of the procedure depends much more on the physician’s ability to perform it. Now, the Food and Drug Administration has approved a new tool that promises to help physicians detect prenatal growth during colonoscopy: a Artificial intelligence The system is made by Medtronic. Physicians say that, among other measures, the tool can help improve the diagnosis. “We really have a chance to completely eradicate colon cancer in anyone who is screened,” said Wallace, who consulted with Medtronic on the project.

The Medtronic system, known as the GI Genius, saw more of the colon inside than most physicians. Medtronic and partner Cosmo Pharmaceuticals reviewed more than 13 million videos of colonoscopy performed in Europe and the United States and trained algorithms to detect polyps that Cosmo collected during drug trials. To “teach” AI to isolate potentially dangerous growths, the images were labeled by gastroenterologists as normal or unhealthy tissue. The AI ​​test was then performed on colonoscopies performed under appropriate conditions and on chronic-to-recognizable polyps, which were only briefly or concealed in a dark spot within the camera range, leading to more difficult challenges.

The system can be added to scopes already used by doctors to perform colonoscopy, with doctors examining the colon and highlighting possible polyps, including a green box. The GI Genius was approved in Europe in October 2012 and is the first AI to be assisted by the FDA in identifying colorectal polyps. “It turned out to be something I missed,” said colleague Wallis. The first validation study GI talent. “It’s an impressive system.”

Mark Pochapin, a gastroenterologist at NYU Langone who is not involved in the creation of GI geniuses, said it was understood that AI polyps would be better to recognize. “There’s less variety when you look at polyps,” Pochapin said. Millions of colonoscopies provide a lot of data to expand the video algorithm. This should protect the system from anxiety Bias in other healthcare algorithms. “There are only so many types of polyps,” he said

Medtronic sees GI Genius and other AI tools as the foundation of its future business, says Giovanni de Napoli, president of Medtronic’s GI business. To that end, the company invested a lot of time and resources to get approval from the FDA for this device. “It took us almost a year to get FDA approval. It’s not easy,” said DA Napoli.

The Medetronic Agency calls its De Novo Pathway under which the FDA sought clearance, requiring applicants to provide information on the safety and effectiveness of the new device, including clinical data. This is a longer and more involved application that has been avoided by some of AII’s other medical devices. Most AI and machine learning medical devices enter the market using a flowing FDA application known as 510 (k) Pathway, which only requires their devices to be proven like other tools already in use and usually takes about six months. According to Published a study The Lancet, Between 2012 and 2020, 92 percent of the 222 AI devices on the market in the United States did so through the A510 (K).



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *