Guiding the Guidance: A Comparative Analysis of User Guidance Signals for Interactive Segmentation of Volumetric Images

TL;DR

This paper compares different user guidance signals for interactive volumetric image segmentation, introduces an adaptive heatmap guidance method, and demonstrates significant performance improvements on medical datasets, advancing clinical applicability.

Contribution

It provides a comprehensive analysis of guidance signals, proposes an adaptive heatmap approach, and shows improved segmentation accuracy in medical imaging tasks.

Findings

Adaptive heatmaps improve segmentation accuracy by 14% Dice.

Choosing the right guidance signal is crucial for interactive segmentation.

The method enhances clinical workflow readiness for medical image segmentation.

Abstract

Interactive segmentation reduces the annotation time of medical images and allows annotators to iteratively refine labels with corrective interactions, such as clicks. While existing interactive models transform clicks into user guidance signals, which are combined with images to form (image, guidance) pairs, the question of how to best represent the guidance has not been fully explored. To address this, we conduct a comparative study of existing guidance signals by training interactive models with different signals and parameter settings to identify crucial parameters for the model's design. Based on our findings, we design a guidance signal that retains the benefits of other signals while addressing their limitations. We propose an adaptive Gaussian heatmaps guidance signal that utilizes the geodesic distance transform to dynamically adapt the radius of each heatmap when encoding clicks. We conduct our study on the MSD Spleen and the AutoPET datasets to explore the segmentation of both anatomy (spleen) and pathology (tumor lesions). Our results show that choosing the guidance signal is crucial for interactive segmentation as we improve the performance by 14% Dice with our adaptive heatmaps on the challenging AutoPET dataset when compared to non-interactive models. This brings interactive models one step closer to deployment on clinical workflows. We will make our code publically available.