Overlap detection and contour-tracking algorithms for critical organs--application to kidney

Area of Science:

• Medical Imaging
• Radiology
• Computer-Aided Diagnosis

Background:

• Accurate segmentation of critical structures in medical imaging is crucial for diagnosis and treatment planning.
• Computer Tomographic (CT) imaging presents challenges in auto-contouring due to similar CT values between organs and background, especially in overlap regions.

Purpose of the Study:

• To develop and evaluate an auto-contouring technique for critical structures on transverse CT images.
• To address the challenge of similar CT values in overlap regions during automated contour detection.

Main Methods:

• A novel auto-contouring technique identifies and skips regions with similar object and background CT values during contour tracking.
• Discontinuous contours generated are corrected using interpolation methods.
• The method incorporates an overlap detection criterion and can operate with minimal operator intervention.

Main Results:

• The technique successfully segmented kidneys, demonstrating its efficacy in challenging cases.
• The overlap detection criterion ensures accurate tracking in non-overlapping environments, relying solely on CT value thresholds.
• The process requires minimal operator input for initiation.

Conclusions:

• The developed auto-contouring technique offers a robust solution for segmenting critical structures in CT images, particularly in areas with low contrast.
• The method shows promise for improving efficiency and accuracy in medical image analysis.
• Further discussion on the technique's successes and limitations provides insights for future development.