Subtraction-based Stepwise computed tomography post-processing with probabilistically adjusted thresholding for...

Area of Science:

• Radiology
• Medical Imaging
• Interventional Radiology

Background:

• Visualization of iceball boundaries in CT-guided renal cryoablation is limited at the fat-ice interface.
• Frozen perirenal fat exhibits poor contrast with non-frozen fat on CT, hindering intraprocedural monitoring.

Purpose of the Study:

• To evaluate a novel stepwise CT post-processing method for improved iceball boundary demarcation.
• To address the visualization challenges at the fat-ice interface during renal cryoablation.

Main Methods:

• Utilized an in situ fat-muscle phantom for evaluation.
• Implemented a two-step CT post-processing technique: fixed zero-threshold subtraction and kernel density estimation-based threshold subtraction.
• Analyzed pixel-wise fat-attenuation distributions to adjust thresholds.

Main Results:

• The stepwise CT post-processing method significantly improved contrast-to-noise ratio in both fat and non-fat tissues.
• Boundary contrast was selectively enhanced by reducing CT values in non-frozen fat regions and frozen non-fat regions.
• Iceball boundaries visualized with the method aligned accurately with magnetic resonance imaging.

Conclusions:

• The proposed CT post-processing approach offers enhanced iceball demarcation during renal cryoablation.
• This technique shows promise for improving intraprocedural guidance in procedures involving perirenal fat.
• Further clinical validation is warranted to confirm its efficacy in practice.