Artificial intelligence has brought the concept of self-driving cars and personalized ads on the web and many more things and now the reports suggest that it is also invading the world of medicine as well. In radiology, artificial intelligence is increasingly helping the doctors to do their jobs profoundly. A computer program that helps the doctors and experts in diagnosing strokes garnered approval from the United States Food and Administration earlier this year. One of the most intriguing lines of research seeks to train computers to diagnose one of the deadliest of all malignancies, pancreatic cancer when the disease is still readily treatable.
This is the vision of Dr. Elliot Fishman who is a professor of radiology at Johns Hopkins Medicine in Baltimore. Artificial intelligence and radiology are considered to be a natural match as so much of the task of reading images involves pattern recognition. It is a dream that has taken more than decades in making, as stated by Fishman. He was also reported saying that when he started in radiology, they said not to worry about reading the chest x-rays because the computers will read them. That was 35 years ago.
The computers still can’t perform the tasks of reading chest x-rays, despite the sky-high expectations and more than a little hype around the role of artificial intelligence. Fishman is undaunted as he turns this technology on the pancreatic cancer. This disease is a huge challenge. Only about 7% of the patients given a pancreatic cancer diagnosis are alive five years later. One of the reasons the disease is so deadly is that the doctors generally diagnose it when it is too late to remove the tumors with surgical procedures. Fishman and his team are looking for ways to change that by training computers to recognise pancreatic cancer early.
Americans got 40 million CT scans of the abdomen every single year for everything from car accidents to mild back pain. If a computer program with expert abilities could help look for pancreatic tumors in all those scans, imagine how easy it would become. Dr. Karen Horton, chair of the Johns Hopkins radiology department and Fishman’s collaborator on the project was reported saying that this is the ultimate opportunity to be able to diagnose it before you have any symptoms and at a stage where it is even maybe too subtle for a radiologist to be able to detect it.