Neural Annotation Refinement: Development of a New 3D Dataset for Adrenal Gland Analysis

TL;DR

This paper introduces Neural Annotation Refinement (NeAR), a method that improves human annotations of 3D adrenal gland shapes using a learnable implicit function, resulting in better diagnostic models and a new open-source dataset for medical shape analysis.

Contribution

The paper presents NeAR, a novel implicit function-based approach to refine medical annotations and introduces the ALAN dataset for adrenal gland analysis.

Findings

NeAR repairs distorted annotations effectively.

Models trained on NeAR-refined shapes outperform those trained on original annotations.

The ALAN dataset provides a new benchmark for adrenal gland shape analysis.

Abstract

The human annotations are imperfect, especially when produced by junior practitioners. Multi-expert consensus is usually regarded as golden standard, while this annotation protocol is too expensive to implement in many real-world projects. In this study, we propose a method to refine human annotation, named Neural Annotation Refinement (NeAR). It is based on a learnable implicit function, which decodes a latent vector into represented shape. By integrating the appearance as an input of implicit functions, the appearance-aware NeAR fixes the annotation artefacts. Our method is demonstrated on the application of adrenal gland analysis. We first show that the NeAR can repair distorted golden standards on a public adrenal gland segmentation dataset. Besides, we develop a new Adrenal gLand ANalysis (ALAN) dataset with the proposed NeAR, where each case consists of a 3D shape of adrenal gland and its diagnosis label (normal vs. abnormal) assigned by experts. We show that models trained on the shapes repaired by the NeAR can diagnose adrenal glands better than the original ones. The ALAN dataset will be open-source, with 1,584 shapes for adrenal gland diagnosis, which serves as a new benchmark for medical shape analysis. Code and dataset are available at https://github.com/M3DV/NeAR.