Researchers at Washington University in St. Louis are using an emerging medical imaging technique with deep learning to diagnose colorectal cancer in real time.
In a pilot study, investigators developed a deep learning-based pattern recognition optical coherence tomography (OCT) system that automates image processing and identified tumors with 100 percent accuracy, compared with pathology reports.
A paper published in the journal Theranosticscontends that it is the first study to report using OCT combined with deep learning to distinguish healthy colorectal tissue from precancerous polyps and cancerous tissue.
“OCT is an emerging imaging technique to obtain 3-dimensional ‘optical biopsies’ of biological samples with high resolution,” state the study’s authors. “We designed a convolutional neural network to capture the structure patterns in human colon OCT images.”
The convolutional neural network was trained and tested using about 26,000 OCT images acquired from 20 tumor areas, 16 benign areas and six other abnormal areas.
Currently, cancer screening or surveillance for colorectal malignancy is conducted with flexible endoscopy, which involves visual inspection of the mucosal lining of the colon and rectum with an optical camera mounted on the endoscope.
“We think this technology, combined with the colonoscopy endoscope, will be very helpful to surgeons in diagnosing colorectal cancer,” says senior author Quing Zhu, professor of biomedical engineering in the McKelvey School of Engineering at Washington University in St. Louis and professor of radiology at the Mallinckrodt Institute of Radiology at Washington University School of Medicine.
“More research is necessary, but the idea is that when the surgeons use colonoscopy to examine the colon surface, this technology could be zoomed in locally to help make a more accurate diagnosis of deeper precancerous polyps and early-stage cancers vs. normal tissue,” adds Zhu.
Researchers are planning future development of the system as an “optical biopsy” tool to help physicians in real time for early mucosal neoplasms screening and treatment evaluation following initial oncologic therapy.
Originally published on healthdatamanagement.com