Related Experiment Video
Updated: Jul 10, 2025
07:32
Standardized Histomorphometric Evaluation of Osteoarthritis in a Surgical Mouse Model
Published on: May 6, 2020
12.2K
CartiMorph: A framework for automated knee articular cartilage morphometrics
Yongcheng Yao1, Junru Zhong1, Liping Zhang1
1CU Lab of AI in Radiology (CLAIR), Department of Imaging and Interventional Radiology, The Chinese University of Hong Kong, Hong Kong, China.
Medical Image Analysis
|November 22, 2023
Summary
CartiMorph is a new deep learning framework for automated knee articular cartilage analysis. It accurately quantifies cartilage loss and thickness, aiding in osteoarthritis biomarker discovery.
Area of Science:
- Biomedical Imaging
- Medical Image Analysis
- Computational Anatomy
Background:
- Knee articular cartilage degeneration is a hallmark of osteoarthritis.
- Accurate morphometric analysis of cartilage is crucial for understanding disease progression and developing treatments.
- Current methods for cartilage morphometrics can be labor-intensive and lack precision.
Purpose of the Study:
- To introduce CartiMorph, a novel deep learning framework for automated knee articular cartilage morphometrics.
- To enable quantitative assessment of cartilage subregions, including full-thickness cartilage loss (FCL), mean thickness, surface area, and volume.
- To establish robust methods for cartilage analysis, including segmentation, registration, thickness mapping, and parcellation.
Main Methods:
- Development of a deep learning framework (CartiMorph) for hierarchical image feature representation.
- Training and validation of deep learning models for tissue segmentation, template construction, and template-to-image registration.
- Implementation of surface-normal-based thickness mapping, FCL estimation, and rule-based cartilage parcellation.
Main Results:
- CartiMorph demonstrated high accuracy in quantitative metrics, with root-mean-squared deviation for FCL measurements below 8%.
- Strong correlations were observed between CartiMorph and manual segmentation for mean thickness (ρ∈[0.82,0.97]), surface area (ρ∈[0.82,0.98]), and volume (ρ∈[0.89,0.98]).
- The proposed rule-based parcellation method outperformed traditional atlas-based approaches, and FCL measurements showed improved accuracy compared to previous studies.
Conclusions:
- CartiMorph provides a robust and accurate automated framework for knee articular cartilage morphometrics.
- The framework shows significant potential for advancing the discovery of imaging biomarkers for knee osteoarthritis.
- Automated quantitative analysis using CartiMorph can improve the efficiency and precision of cartilage assessment in clinical and research settings.
