SkullEngine: A Multi-stage CNN Framework for Collaborative CBCT Image Segmentation and Landmark Detection

TL;DR

SkullEngine is a multi-stage CNN framework designed for precise CBCT image segmentation and landmark detection, utilizing a collaborative JSD model and refinement stages to improve accuracy on clinical datasets.

Contribution

It introduces a novel multi-stage CNN approach with collaborative models for high-resolution segmentation and landmark detection in CBCT images.

Findings

Effective segmentation of midface and mandible bones.

Accurate detection of 175 clinical landmarks.

Validated on 170 CBCT/CT images.

Abstract

We propose a multi-stage coarse-to-fine CNN-based framework, called SkullEngine, for high-resolution segmentation and large-scale landmark detection through a collaborative, integrated, and scalable JSD model and three segmentation and landmark detection refinement models. We evaluated our framework on a clinical dataset consisting of 170 CBCT/CT images for the task of segmenting 2 bones (midface and mandible) and detecting 175 clinically common landmarks on bones, teeth, and soft tissues.