SU-E-J-111: Compute the Optimal Threshold for PET Functional Volume Delineation

Area of Science:

• Medical Imaging
• Image Analysis
• Quantitative PET

Background:

• Partial Volume Effect (PVE) in PET imaging complicates accurate volume delineation.
• Existing methods often require scanner-specific parameters or tumor size information.
• Functional volume delineation is crucial for accurate disease assessment.

Purpose of the Study:

• To develop a method for computing the optimal threshold (OT) directly from PET images.
• To enable functional volume delineation without a priori information on scanners, protocols, or tumor size.

Main Methods:

• Analyzed a mathematical model for PVE in spherical objects with Gaussian Point Spread Function (PSF).
• Introduced the Optimal Area Ratio (OAR) concept to recover PSF and tumor size information from PET images.
• Utilized the dynamic behavior of a region-growing algorithm across thresholds [0,1] to determine OT.
• Validated the method through computer simulations and phantom experiments with varying object sizes and noise levels.

Main Results:

• Computer simulations demonstrated exact calculation of the real OT.
• Phantom experiments yielded visually satisfactory functional volume delineation using the calculated OT.
• The method successfully recovered embedded information about PSF and tumor size.

Conclusions:

• The developed method allows for OT calculation directly from PET images, independent of external information.
• Theoretical analysis and experimental validation confirm the efficacy of the approach for PET functional volume delineation.
• This method simplifies and enhances the accuracy of PET-based volume measurements.