Decoupled Gradient Harmonized Detector for Partial Annotation: Application to Signet Ring Cell Detection

TL;DR

This paper introduces a novel loss function, DGHM-C, that improves detection of signet ring cells in medical images with partial annotations by decoupling noisy and clean examples and harmonizing their gradient distributions.

Contribution

The paper proposes the Decoupled Gradient Harmonizing Mechanism (DGHM) integrated into the classification loss to handle partial annotations and label noise in medical image detection tasks.

Findings

Achieved 2nd place in MICCAI DigestPath2019 challenge.

DGHM-C loss significantly improves detection accuracy with partial annotations.

Demonstrated robustness to label missing rates through experiments.

Abstract

Early diagnosis of signet ring cell carcinoma dramatically improves the survival rate of patients. Due to lack of public dataset and expert-level annotations, automatic detection on signet ring cell (SRC) has not been thoroughly investigated. In MICCAI DigestPath2019 challenge, apart from foreground (SRC region)-background (normal tissue area) class imbalance, SRCs are partially annotated due to costly medical image annotation, which introduces extra label noise. To address the issues simultaneously, we propose Decoupled Gradient Harmonizing Mechanism (DGHM) and embed it into classification loss, denoted as DGHM-C loss. Specifically, besides positive (SRCs) and negative (normal tissues) examples, we further decouple noisy examples from clean examples and harmonize the corresponding gradient distributions in classification respectively. Without whistles and bells, we achieved the 2nd place in the challenge. Ablation studies and controlled label missing rate experiments demonstrate that DGHM-C loss can bring substantial improvement in partially annotated object detection.