Early esophageal cancer detection using RF classifiers

摘要: Esophageal cancer is one of the fastest rising forms in Western world. Using High-Definition (HD) endoscopy, gastroenterology experts can identify esophageal at an early stage. Recent research shows that be found using a state-of-the-art computer-aided detection (CADe) system based on analyzing static HD endoscopic images. Our aims extending this by applying Random Forest (RF) classification, which introduces confidence measure for detected regions. To visualize data, we propose novel automated annotation system, employing unique characteristics previous measure. This approach allows reliable modeling multi-expert knowledge and provides essential data real-time video processing, to enable future use clinical setting. The performance CADe evaluated 39-patient dataset, containing 100 images annotated 5 expert gastroenterologists. proposed reaches precision 75% recall 90%, thereby improving results 11 6 percentage points, respectively.