One-Shot Medical Landmark Detection

TL;DR

This paper introduces CC2D, a novel one-shot learning framework for medical landmark detection that effectively reduces the need for large annotated datasets by leveraging self-supervised learning and pseudo-label training.

Contribution

The paper presents a new two-stage framework, CC2D, for one-shot landmark detection that achieves competitive accuracy with only a single annotated image, reducing data annotation burdens.

Findings

Achieves 81.01% detection accuracy within 4.0mm on public dataset.

Comparable to state-of-the-art fully-supervised methods with much less data.

Demonstrates effectiveness of self-supervised and pseudo-label strategies in medical imaging.

Abstract

The success of deep learning methods relies on the availability of a large number of datasets with annotations; however, curating such datasets is burdensome, especially for medical images. To relieve such a burden for a landmark detection task, we explore the feasibility of using only a single annotated image and propose a novel framework named Cascade Comparing to Detect (CC2D) for one-shot landmark detection. CC2D consists of two stages: 1) Self-supervised learning (CC2D-SSL) and 2) Training with pseudo-labels (CC2D-TPL). CC2D-SSL captures the consistent anatomical information in a coarse-to-fine fashion by comparing the cascade feature representations and generates predictions on the training set. CC2D-TPL further improves the performance by training a new landmark detector with those predictions. The effectiveness of CC2D is evaluated on a widely-used public dataset of cephalometric landmark detection, which achieves a competitive detection accuracy of 81.01\% within 4.0mm, comparable to the state-of-the-art fully-supervised methods using a lot more than one training image.