Development and Evaluation of Urolithiasis Detection Technology Based on a Multimethod Algorithm

Area of Science:

• Medical Imaging
• Artificial Intelligence in Medicine
• Urology

Background:

• Accurate detection of ureter stones is crucial for effective treatment planning.
• Current diagnostic methods may have limitations in sensitivity and specificity.
• The integration of artificial intelligence offers potential for improved diagnostic accuracy.

Purpose of the Study:

• To propose an optimal ureter stone detection model using multiple artificial intelligence technologies.
• To develop a multimethod approach by merging artificial intelligence and image processing models for urinary tract stone detection.
• To enhance the accuracy of urinary tract stone detection.

Main Methods:

• Utilizing artificial intelligence technology for an optimal urinary tract stone detection algorithm.
• Combining deep learning technology (Fast R-CNN) with image processing technology (Watershed).
• Developing a multimethod approach for enhanced stone detection.

Main Results:

• Achieved a sensitivity of 0.90 and specificity of 0.91 for urinary tract stone detection.
• Obtained a positional accuracy of 0.84, exceeding the standard accuracy threshold of 0.8.
• Demonstrated the platform's capability to provide accurate guidance for surgical interventions.

Conclusions:

• The proposed method effectively assists in diagnostic decision-making within clinically acceptable safety margins.
• The combined diagnostic approach shows significant value, particularly for complex cases like ureter stones with polyps.
• The AI-powered model supports therapeutic decisions and improves patient outcomes in urological care.