PnPNet: Pull-and-Push Networks for Volumetric Segmentation with Boundary Confusion

TL;DR

This paper introduces PnPNet, a novel volumetric segmentation network that employs push-and-pull mechanisms to improve boundary delineation, especially in boundary-confused regions, outperforming existing methods on multiple datasets.

Contribution

The paper proposes PnPNet with push and pull branches to model boundary shape characteristics, addressing boundary confusion more effectively than prior U-shape networks.

Findings

PnPNet outperforms other segmentation networks on HD and ASSD metrics.

Push and pull modules can enhance classic U-shape models as plug-and-play components.

The method effectively handles boundary confusion in volumetric images.

Abstract

Precise boundary segmentation of volumetric images is a critical task for image-guided diagnosis and computer-assisted intervention, especially for boundary confusion in clinical practice. However, U-shape networks cannot effectively resolve this challenge due to the lack of boundary shape constraints. Besides, existing methods of refining boundaries overemphasize the slender structure, which results in the overfitting phenomenon due to networks' limited abilities to model tiny objects. In this paper, we reconceptualize the mechanism of boundary generation by encompassing the interaction dynamics with adjacent regions. Moreover, we propose a unified network termed PnPNet to model shape characteristics of the confused boundary region. Core ingredients of PnPNet contain the pushing and pulling branches. Specifically, based on diffusion theory, we devise the semantic difference module (SDM) from the pushing branch to squeeze the boundary region. Explicit and implicit differential information inside SDM significantly boost representation abilities for inter-class boundaries. Additionally, motivated by the K-means algorithm, the class clustering module (CCM) from the pulling branch is introduced to stretch the intersected boundary region. Thus, pushing and pulling branches will shrink and enlarge the boundary uncertainty respectively. They furnish two adversarial forces to promote models to output a more precise delineation of boundaries. We carry out experiments on three challenging public datasets and one in-house dataset, containing three types of boundary confusion in model predictions. Experimental results demonstrate the superiority of PnPNet over other segmentation networks, especially on evaluation metrics of HD and ASSD. Besides, pushing and pulling branches can serve as plug-and-play modules to enhance classic U-shape baseline models. Codes are available.