A material sensitivity study on the accuracy of deformable organ registration using linear biomechanical models

Area of Science:

• Medical Physics
• Computational Biology
• Biomechanical Engineering

Background:

• Model-based deformable organ registration using finite element method (FEM) is crucial for image-guided adaptive radiotherapy (IGART).
• Accurate measurement of patient-specific tissue mechanical properties is challenging, impacting registration accuracy.
• Existing techniques often assume linear elasticity and predetermined material properties.

Purpose of the Study:

• To systematically investigate the relationship between achievable registration accuracy and tissue mechanical/organ geometrical properties.
• To establish the sensitivity of registration errors to variations in tissue elastic material constants.
• To evaluate registration accuracy for rectal wall, bladder wall, and prostate under material uncertainty.

Main Methods:

• Development of biomechanical models for rectal wall, bladder wall, and prostate using simplified and patient-specific geometries.
• Implementation of orthotropic or transversely isotropic elastic models to capture tissue anisotropy.
• Computer simulations of organ deformation and image registration with perturbed material properties.

Main Results:

• Registration error increases with distance from the organ boundary.
• Material stability is a dominant factor influencing registration accuracy under uncertainty.
• For hollow organs (rectum, bladder) with 30% material uncertainty, errors are limited to 1.3 mm.
• For solid organs (prostate) with 30% material uncertainty, errors can reach 4.5 mm, though most subvolumes show smaller errors.

Conclusions:

• Deformable organ registration using FEM is viable for IGART if mean material parameters are available.
• Registration accuracy is significantly affected by material property uncertainty, particularly for solid organs.
• Understanding the interplay between biomechanics and geometry is key to improving registration precision.