RECIST-Net: Lesion detection via grouping keypoints on RECIST-based annotation

TL;DR

RECIST-Net introduces a novel lesion detection method in CT images that detects keypoints based on RECIST diameters, simplifying the process and improving sensitivity over existing bounding-box methods.

Contribution

It proposes a new keypoint detection approach for RECIST-based lesion detection, avoiding complex anchor design and shape loss issues of traditional methods.

Findings

Achieves 92.49% sensitivity at four false positives per image.

Outperforms recent multi-task learning methods.

Provides a conceptually straightforward detection formulation.

Abstract

Universal lesion detection in computed tomography (CT) images is an important yet challenging task due to the large variations in lesion type, size, shape, and appearance. Considering that data in clinical routine (such as the DeepLesion dataset) are usually annotated with a long and a short diameter according to the standard of Response Evaluation Criteria in Solid Tumors (RECIST) diameters, we propose RECIST-Net, a new approach to lesion detection in which the four extreme points and center point of the RECIST diameters are detected. By detecting a lesion as keypoints, we provide a more conceptually straightforward formulation for detection, and overcome several drawbacks (e.g., requiring extensive effort in designing data-appropriate anchors and losing shape information) of existing bounding-box-based methods while exploring a single-task, one-stage approach compared to other RECIST-based approaches. Experiments show that RECIST-Net achieves a sensitivity of 92.49% at four false positives per image, outperforming other recent methods including those using multi-task learning.