Transfer Learning for Effective Urolithiasis Detection.

Area of Science:

• Artificial Intelligence
• Medical Imaging
• Urology

Background:

• Urolithiasis (urinary tract stones) is a prevalent condition causing significant pain and potential complications.
• Accurate and timely diagnosis is crucial for effective patient management and treatment.

Purpose of the Study:

• To develop a deep learning model for rapid and accurate detection of urinary tract stones.
• To leverage transfer learning for enhanced diagnostic performance.
• To improve the efficiency of medical staff in diagnosing urolithiasis.

Main Methods:

• Utilized the ResNet50 architecture for feature extraction.
• Applied transfer learning by fine-tuning pretrained model weights.
• Evaluated model performance using accuracy, precision-recall, and ROC curve analysis.

Main Results:

• The ResNet50-based deep learning model demonstrated high accuracy and sensitivity in detecting urinary tract stones.
• The model significantly outperformed traditional diagnostic methods.
• Enabled rapid diagnosis, assisting clinicians in decision-making.

Conclusions:

• The developed deep learning model accelerates the clinical application of urinary tract stone detection technology.
• This AI tool enhances medical staff efficiency through swift identification of stones.
• Contributes to the advancement of deep learning-based medical imaging diagnostics.