At the Eidhoven University of Technology (TU Eindhoven) in The Netherlands, a research team has developed a computer vision system that has shown excellent results at identifying early neoplastic lesions, which develop into full blown esophageal cancer, in patients with Barrett’s esophagus. Such lesions are very difficult to spot, and not many physicians have the training or the eye necessary to do so accurately and consistently.
The collaboration involved a gastroenterologist from Catharina Hospital and computer scientists from the Video Coding and Architectures Research Group at TU Eindhoven. Their algorithm analyzes endoscopy images of Barrett’s esophagus, recognizing slight differences in the texture and color of the tissue observed. This is compared to data gathered from previously analyzed images of Barrett’s patients with and without neoplastic lesions. The computer learning algorithm has got acquainted over time to noticing the minute fluctuations in the image and in their study, comparing it to four international experts in the field, achieved a nearly perfect score.
From the abstract in journal Endoscopy:
The system identified early neoplastic lesions on a per-image analysis with a sensitivity and specificity of 0.83. At the patient level, the system achieved a sensitivity and specificity of 0.86 and 0.87, respectively. A trade-off between the two performance metrics could be made by varying the percentage of training samples that showed neoplastic tissue.
The automated computer algorithm developed in this study was able to identify early neoplastic lesions with reasonable accuracy, suggesting that automated detection of early neoplasia in Barrett’s esophagus is feasible. Further research is required to improve the accuracy of the system and prepare it for real-time operation, before it can be applied in clinical practice.
Study in journal Endoscopy: Computer-aided detection of early neoplastic lesions in Barrett’s esophagus…
Via: TU Eindhoven…