A Robust Rule-Based Framework for Stone Detection and Posterior Acoustic Shadow Localization in Abdominal Ultrasound

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Computational Pathology

Background:

• Posterior acoustic shadowing is a known ultrasound artifact linked to calcified stones.
• Current automated ultrasound analysis methods have not fully utilized this phenomenon.
• Accurate stone detection and localization are crucial for clinical management.

Purpose of the Study:

• To develop an explainable, semi-automatic framework for stone detection and localization.
• To incorporate posterior acoustic shadow characteristics into the analysis.
• To enhance diagnostic confidence and reproducibility in ultrasound-based stone assessment.

Main Methods:

• A rule-based framework was designed using morphological enhancement to identify stone candidates.
• Candidates were evaluated using local contrast analysis, posterior shadow assessment, and shape penalties.
• A composite score integrated these features for candidate ranking.
• The framework was tested on kidney and gallbladder stone ultrasound images.

Main Results:

• In the 50-60% confidence group, accuracy for kidney stones rose from 0.29 to 0.64 and for gallbladder stones from 0.30 to 0.60 with posterior shadow inclusion.
• Centroid distance errors in the ≥80% confidence group were 1.26 ± 0.28 mm (kidney stones) and 1.44 ± 0.91 mm (gallbladder stones).
• The framework demonstrated improved diagnostic confidence and accuracy.

Conclusions:

• The proposed framework effectively leverages posterior acoustic shadow analysis for improved stone detection.
• This method offers a reproducible augmentation to conventional ultrasound assessments.
• The explainable, rule-based approach enhances clinical utility in stone evaluation.