Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ConUNETR: A Conditional Transformer Network for 3D Micro-CT Embryonic Cartilage Segmentation

Nishchal Sapkota, Yejia Zhang, Susan M. Motch Perrine, Yuhan Hsi, Sirui Li, Meng Wu, Greg Holmes, Abdul R. Abdulai, Ethylin W. Jabs, Joan T. Richtsmeier, Danny Z Chen

TL;DR

The paper tackles embryonic cartilage segmentation across rapid developmental changes where morphology varies by age and mutation. It introduces ConUNETR, a lightweight Transformer-based segmentation model conditioned on age priors via learnable age tokens and spatial encoding integrated into a ViT encoder. The approach yields improved generalization across unseen ages and mutation cohorts, with ablation showing the positive impact of spatial encoding and conditional components. This work offers a practical path toward efficient, scalable cartilage segmentation in developmental studies, reducing annotation burden and enabling cross-age analyses.

Abstract

Studying the morphological development of cartilaginous and osseous structures is critical to the early detection of life-threatening skeletal dysmorphology. Embryonic cartilage undergoes rapid structural changes within hours, introducing biological variations and morphological shifts that limit the generalization of deep learning-based segmentation models that infer across multiple embryonic age groups. Obtaining individual models for each age group is expensive and less effective, while direct transfer (predicting an age unseen during training) suffers a potential performance drop due to morphological shifts. We propose a novel Transformer-based segmentation model with improved biological priors that better distills morphologically diverse information through conditional mechanisms. This enables a single model to accurately predict cartilage across multiple age groups. Experiments on the mice cartilage dataset show the superiority of our new model compared to other competitive segmentation models. Additional studies on a separate mice cartilage dataset with a distinct mutation show that our model generalizes well and effectively captures age-based cartilage morphology patterns.

ConUNETR: A Conditional Transformer Network for 3D Micro-CT Embryonic Cartilage Segmentation

TL;DR

The paper tackles embryonic cartilage segmentation across rapid developmental changes where morphology varies by age and mutation. It introduces ConUNETR, a lightweight Transformer-based segmentation model conditioned on age priors via learnable age tokens and spatial encoding integrated into a ViT encoder. The approach yields improved generalization across unseen ages and mutation cohorts, with ablation showing the positive impact of spatial encoding and conditional components. This work offers a practical path toward efficient, scalable cartilage segmentation in developmental studies, reducing annotation burden and enabling cross-age analyses.

Abstract

Studying the morphological development of cartilaginous and osseous structures is critical to the early detection of life-threatening skeletal dysmorphology. Embryonic cartilage undergoes rapid structural changes within hours, introducing biological variations and morphological shifts that limit the generalization of deep learning-based segmentation models that infer across multiple embryonic age groups. Obtaining individual models for each age group is expensive and less effective, while direct transfer (predicting an age unseen during training) suffers a potential performance drop due to morphological shifts. We propose a novel Transformer-based segmentation model with improved biological priors that better distills morphologically diverse information through conditional mechanisms. This enables a single model to accurately predict cartilage across multiple age groups. Experiments on the mice cartilage dataset show the superiority of our new model compared to other competitive segmentation models. Additional studies on a separate mice cartilage dataset with a distinct mutation show that our model generalizes well and effectively captures age-based cartilage morphology patterns.
Paper Structure (10 sections, 3 equations, 1 figure, 3 tables)

This paper contains 10 sections, 3 equations, 1 figure, 3 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Architecture
  4. Age Tokens
  5. Spatial Encoding
  6. Datasets and Implementation Details
  7. Ablation Study
  8. Conclusions
  9. COMPLIANCE WITH ETHICAL STANDARDS
  10. Acknowledgments

Figures (1)

  • Figure 1: An overview of the architecture of our proposed Conditional Universal Model (ConUNETR).