Attention residual U-Net improves polyp detection across endoscopy devices
Attention Residual U-Net for Enhanced Cross-Dataset Polyp Localization in Endoscopy
Computer Vision and Pattern Recognition
Summary
Detecting small or unclear polyps in endoscopy images can be hard because images vary a lot depending on the device and situation. The authors created a new deep learning method that focuses more accurately on the important parts of the images, ignoring irrelevant details. This approach helps locate polyps better across different types of endoscopy images. Tests showed it works better than many current methods, improving detection accuracy by a notable margin.
What this means in practice
- •For medical imaging developers: Enhance software that assists doctors by improving polyp location accuracy in endoscopic images from various devices.
- •For healthcare ai solution providers: Create more robust AI tools for detecting intestinal lesions that maintain performance despite varying clinical imaging conditions.$Commercial implications: This method enables commercially viable AI products offering improved polyp detection across diverse endoscopy hardware and patient conditions.
Authors
Romil Imtiaz, Dimitris K. Iakovidis
Abstract
Precise localization of pathologies in endoscopy is a critical task in medical imaging, with deep learning (DL) frameworks now serving as the primary methodology for automated lesion detection. Today, generic object detection methods, such as YOLO, provide state-of-the-art performance in this context. However, their robustness often decreases under real-world clinical conditions due to variations across endoscopic devices, intestinal content, dataset bias, and the presence of small or low-contrast lesions. Motivated by these challenges, we propose a novel polyp detection framework based on a Attention Residual U-Net offering enhanced polyp localization. Instead of directly regressing bounding boxes from global feature representations, the proposed method learns lesion-focused spatial representations and converts them into bounding-box predictions during inference. Unlike conventional encoder-decoder localization frameworks that directly propagate all encoder activations into the decoding pathway, the proposed method selectively suppresses irrelevant background responses before feature reconstruction. This encourages the network to focus on lesion-relevant spatial evidence under heterogeneous imaging conditions. To achieve this, the framework integrates residual feature propagation, attention-guided encoder filtering, and multi-scale contextual learning. The proposed framework was evaluated on multiple public endoscopic datasets acquired using different devices. The comparative experimental results validate its effectiveness over the relevant state-of-the-art methods, with an improvement up to 22.0% in terms of the area under the receiving operating characteristic (AUC).