Artificial Intelligence (AI) can accurately detect or diagnose colorectal cancer from tissue scans, probably better than pathologists, according to a study conducted by the researchers of Tulane University and other universities.
The study, published in the journal Nature Communications, was designed to test whether AI could help pathologists keep pace with the rising demand for their services, per Science Daily.
Pathologists analyze, evaluate and label thousands of histopathology images regularly to determine whether someone has cancer. However, their workload has increased significantly, which may sometimes cause unintended misdiagnoses due to stress or fatigue.
Study author Dr. Hong-Wen Deng of Tulane University said, “Even though a lot of their work is repetitive, most pathologists are extremely busy because there’s a huge demand for what they do but there’s a global shortage of qualified pathologists, especially in many developing countries.”
“This study is revolutionary because we successfully leveraged artificial intelligence to identify and diagnose colorectal cancer in a cost-effective way,” he added, “which could ultimately reduce the workload of pathologists.”
Dr. Deng and his colleagues looked at more than 13,000 images of colorectal cancer from over 8,800 subjects and 13 independent cancer centers in China, Germany, and the United States. They built a machine-assisted pathological recognition program, which allows a computer to recognize images that show colorectal cancer.
Dr. Deng said, “The challenges of this study stemmed from complex large image sizes, complex shapes, textures, and histological changes in nuclear staining. But ultimately the study revealed that when we used AI to diagnose colorectal cancer, the performance is shown comparable to and even better in many cases than real pathologists.”
Upon comparing the computer’s results with highly experienced pathologists who interpreted data manually, the researchers found that the average pathologist accurately identified colorectal cancer manually.
Using AI to identify cancer is one of the emerging technologies, but it has not yet been widely accepted.
Dr. Deng hopes that the study findings will lead to more pathologists using prescreening technology to make quicker diagnoses.
“It’s still in the research phase and we haven’t commercialized it yet because we need to make it more user friendly and test and implement in more clinical settings,” he said. “But as we develop it further, hopefully, it can also be used for different types of cancer in the future.”
“Using AI to diagnose cancer can expedite the whole process and will save a lot of time for both patients and clinicians,” explained Dr. Deng.