Design of a predictive targeting error simulator for MRI-guided prostate biopsy

Area of Science:

• Medical Imaging
• Computational Biology
• Surgical Simulation

Background:

• Multi-parametric MRI offers superior quality compared to Ultrasound (US) for prostate imaging.
• Current standard prostate biopsy procedures utilize US, facing limitations in accuracy.
• Surface-based registration is a simpler alternative to deformable registration for aligning pre-operative and intra-operative prostate volumes.

Purpose of the Study:

• To introduce and detail a novel, interactive simulator for validating prostate biopsy targeting accuracy.
• To quantify the impact of segmentation errors on registration and biopsy targeting.
• To establish preliminary tolerable limits for segmentation errors in the context of small tumor detection.

Main Methods:

• Development of a fully interactive and customizable simulator for prostate biopsy registration.
• Implementation of a segmentation error introduction algorithm using polar curves and natural cubic spline interpolation.
• Simulation of organ deformation, segmentation, and targeting with user-defined parameters.

Main Results:

• The simulator quantifies targeting errors resulting from simulated registrations between prostate volumes.
• Statistically realistic contouring errors are introduced via a novel algorithm.
• Preliminary simulations indicate a maximum tolerable segmentation error of 4-5 mm for a 5.0 mm small tumor.

Conclusions:

• The developed simulator provides a robust platform for validating prostate biopsy procedures.
• Segmentation accuracy is critical for accurate registration and biopsy targeting.
• Further validation through clinical trials is planned to confirm simulation findings.