Research to explore if AI could be used to assist pathologists in keeping up with the increased demand for medical services.
Pathologists regularly review and categorize thousands of histopathology scans to determine if somebody has cancer. However, their typical workload has increased dramatically, which can often lead to unintentional misdiagnosis owing to exhaustion. Despite the fact that much of their job is repetitive, most pathologists are quite busy since there is a high demand for their services, yet there is a global scarcity of skilled pathologists, particularly in many developing countries. This study is groundbreaking since the researchers used ai technology to discover and diagnose colorectal cancer in a cost-effective manner, potentially reducing pathologists' workload.
The team collected around 13,000 colorectal cancer photos from 8,803 people and 13 independent cancer institutions in the U.S., China, and Germany for the study. Using photographs chosen at random by technicians, they created a machine-assisted pathological recognition program that enables a computer to recognize images of colorectal cancer, which is one of the leading causes of cancer-related deaths in the U.S. and Europe.
The difficulties in this research resulted from the enormous input images, complicated shapes and textures, and histological variations in nuclear staining. However, the study concluded that when they utilized AI to identify colorectal cancer, the performance was equivalent to, if not better than, that of real pathologists in many cases. The researchers employed the area under the receiver operating characteristic curve, or AUC, as a performance measuring method to estimate the study's success. When the computer's results were compared against the work of highly skilled pathologists who manually read data, the study discovered that the typical pathologist scored .969 for correctly classifying colorectal cancer. The average score for the machine-assisted AI computer program was .98, which is comparable to, if not more accurate than, human intelligence.
Using AI technology to detect cancer is a new technology that has not yet gained widespread acceptance. The team expects that the research will lead to more pathologists adopting prescreening technologies to make faster diagnoses in the future.
