Statistical Biomechanical Surface Registration: Application to MR-TRUS Fusion for Prostate Interventions

PROJECT TITLE :

Statistical Biomechanical Surface Registration: Application to MR-TRUS Fusion for Prostate Interventions

ABSTRACT:

A common challenge when performing surface-primarily based registration of images is ensuring that the surfaces accurately represent consistent anatomical boundaries. Image segmentation could be tough in some regions due to either poor distinction, low slice resolution, or tissue ambiguities. To address this, we present a unique non-rigid surface registration methodology designed to register 2 partial surfaces, capable of ignoring regions where the anatomical boundary is unclear. Our probabilistic approach incorporates previous geometric info in the shape of a statistical shape model (SSM), and physical knowledge in the shape of a finite part model (FEM). We validate results in the context of prostate interventions by registering pre-operative magnetic resonance imaging (MRI) to 3D transrectal ultrasound (TRUS). We show that both the geometric and physical priors significantly decrease internet target registration error (TRE), leading to TREs of two.35 $pm$ zero.81 mm and 2.eighty one $pm$ zero.sixty six mm when applied to full and partial surfaces, respectively. We have a tendency to investigate robustness in response to errors in segmentation, varying levels of missing data, and adjusting the tunable parameters. Results demonstrate that the proposed surface registration methodology is an efficient, sturdy, and effective resolution for fusing knowledge from multiple modalities.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here