Automated pancreas segmentation from three-dimensional contrast-enhanced computed tomography.

Area of Science:

• Medical Imaging
• Computer-Aided Diagnosis
• Radiology

Background:

• Pancreas segmentation is critical for diagnosing and monitoring pancreatic diseases.
• Manual segmentation of the pancreas on computed tomography (CT) scans is time-consuming and prone to inter-observer variability.
• Automated segmentation methods are needed to improve efficiency and consistency in clinical practice.

Purpose of the Study:

• To develop and validate an automated algorithm for pancreas segmentation from contrast-enhanced multiphase CT.
• To assess the effectiveness of the proposed algorithm in accurately delineating the pancreas region.

Main Methods:

• A two-stage segmentation strategy with spatial standardization was employed to address variations in pancreas shape and location.
• Patient-specific probabilistic atlases were utilized to guide segmentation and manage residual variability.
• A classifier ensemble was incorporated to refine initial segmentation results and enhance accuracy.

Main Results:

• The algorithm was validated on 20 unknown CT volumes and 3 competition datasets.
• The proposed algorithm demonstrated enhanced segmentation performance.
• The Jaccard index achieved between the segmented and true pancreas was 57.9%.

Conclusions:

• The study confirmed the effectiveness of the two-stage segmentation with spatial standardization for pancreas delineation.
• Patient-specific probabilistic atlas guided segmentation proved effective in reducing false negatives.
• The classifier ensemble successfully boosted overall segmentation performance.