Statistical multi-level shape models for scalable modeling of multi-organ anatomies

Area of Science:

• Medical imaging analysis
• Computational anatomy
• Statistical modeling

Background:

• Statistical shape modeling (SSM) is crucial for quantitative anatomical analysis.
• Particle-based shape modeling (PSM) learns population-level shape from medical imaging.
• Current global multi-organ models lack scalability and induce anatomical inconsistencies.

Purpose of the Study:

• To develop an efficient multi-organ modeling approach.
• To capture inter-organ relations and optimize morphological changes simultaneously.
• To overcome limitations of global statistical models in multi-organ analysis.

Main Methods:

• Leveraging Particle-based Shape Modeling (PSM).
• Proposing a new correspondence-point optimization for multiple organs.
• Utilizing multilevel component analysis to create orthogonal within- and between-organ subspaces.
• Formulating the correspondence optimization objective using a generative model.

Main Results:

• The proposed method overcomes limitations of global multi-organ models.
• Achieves scalable and consistent multi-organ shape statistics.
• Effectively captures inter-organ pose variations and individual organ morphology.
• Demonstrated efficacy on synthetic and clinical data (spine, foot/ankle, hip).

Conclusions:

• The multilevel component analysis provides a robust framework for multi-organ statistical shape modeling.
• This approach enhances the accuracy and interpretability of shape variations in complex anatomical structures.
• The method offers significant improvements for analyzing population-level shape statistics of articulated joints.