PDSE: A Multiple Lesion Detector for CT Images using PANet and Deformable Squeeze-and-Excitation Block

TL;DR

This paper presents PDSE, a one-stage lesion detection framework for CT images that improves accuracy and efficiency by enhancing feature aggregation and incorporating attention mechanisms, achieving state-of-the-art results on the DeepLesion benchmark.

Contribution

The paper introduces PDSE, a novel lesion detection model that redesigns Retinanet with feature enhancement and attention modules for better detection of diverse lesions.

Findings

Achieved over 0.20 mAP on DeepLesion benchmark.

Significantly improved detection of small and multiscaled lesions.

Outperformed existing advanced algorithms in lesion detection accuracy.

Abstract

Detecting lesions in Computed Tomography (CT) scans is a challenging task in medical image processing due to the diverse types, sizes, and locations of lesions. Recently, various one-stage and two-stage framework networks have been developed to focus on lesion localization. We introduce a one-stage lesion detection framework, PDSE, by redesigning Retinanet to achieve higher accuracy and efficiency for detecting lesions in multimodal CT images. Specifically, we enhance the path aggregation flow by incorporating a low-level feature map. Additionally, to improve model representation, we utilize the adaptive Squeeze-and-Excitation (SE) block and integrate channel feature map attention. This approach has resulted in achieving new state-of-the-art performance. Our method significantly improves the detection of small and multiscaled objects. When evaluated against other advanced algorithms on the public DeepLesion benchmark, our algorithm achieved an mAP of over 0.20.