Heatmap Regression for Lesion Detection using Pointwise Annotations

TL;DR

This paper introduces a heatmap regression-based lesion detection method that uses only point annotations, enabling effective detection and uncertainty estimation without requiring detailed segmentation labels, and facilitates efficient transfer to segmentation tasks.

Contribution

The novel approach leverages point labels for lesion detection with probabilistic outputs and uncertainty estimation, reducing annotation effort and improving transfer learning for segmentation.

Findings

Competitive detection performance with point labels

Reliable lesion existence uncertainty estimation

Effective pre-training for segmentation with limited data

Abstract

In many clinical contexts, detecting all lesions is imperative for evaluating disease activity. Standard approaches pose lesion detection as a segmentation problem despite the time-consuming nature of acquiring segmentation labels. In this paper, we present a lesion detection method which relies only on point labels. Our model, which is trained via heatmap regression, can detect a variable number of lesions in a probabilistic manner. In fact, our proposed post-processing method offers a reliable way of directly estimating the lesion existence uncertainty. Experimental results on Gad lesion detection show our point-based method performs competitively compared to training on expensive segmentation labels. Finally, our detection model provides a suitable pre-training for segmentation. When fine-tuning on only 17 segmentation samples, we achieve comparable performance to training with the full dataset.