An artificial intelligence (AI) tool is very effective at detecting pancreatic cancer on CT, according to a study published in Radiology, the journal of the Radiological Society of North America (RSNA).
Pancreatic cancer has the lowest five-year survival rate among cancers. It is projected to be the second leading cause of cancer death in the United States by 2030.
Early detection is the best way to improve the prognosis, as the prognosis worsens if the tumor grows more than 2 centimeters. CT is the most important method for detecting pancreatic cancer, but it misses about 40% of tumors smaller than 2 centimeters.
There is an urgent need for an effective tool to help radiologists improve pancreatic cancer detection. Researchers in Taiwan studied a computer-aided detection (CAD) tool that used a type of artificial intelligence called deep learning to detect pancreatic cancer.
Pancreatic cancer from non-cancerous pancreas. But that study relied on radiologists manually identifying the pancreas on imaging — a labor-intensive process known as segmentation. In the new study, an AI tool automatically identified the pancreas. This is an important development because the pancreas borders many organs and structures and varies in shape and size.
The researchers developed the tool with an internal test set consisting of 546 pancreatic cancer patients and 733 control participants. The tool achieved 90% sensitivity and 96% specificity in an internal test set.
Validation was followed by a set of 1,473 individual CT scans from institutions across Taiwan. The tool achieved 90% sensitivity and 93% specificity in identifying pancreatic cancer from controls in this cohort. The sensitivity for detecting pancreatic cancer smaller than 2 centimeters is 75%.
"The performance of the deep learning tool appears to be comparable to radiologists," said lead study author Weichung Wang, Ph.D., a professor at National Taiwan University and director of the university's MeDA Lab. "Notably, in this study, the sensitivity of a computer-assisted deep learning detection tool for pancreatic cancer was comparable to that of radiologists at a tertiary referral center, regardless of tumor size and stage."
The CAD tool has the potential to provide a lot of information to help clients, says Dr. Wang said. This may indicate an area of suspicion to facilitate interpretation by the radiologist.
"The CAD tool can serve as an adjunct for radiologists to improve pancreatic cancer detection," said study co-author Wei-Chi Liao, M.D., Ph.D., of National Taiwan University and National Taiwan University Hospital. . The researchers are planning further studies. In particular, they want to look at the performance of the tool in a more diverse population.
Since the current study was retrospective, they wanted to see how it would perform in the future in a real-world clinical setting.