Related Experiment Video
Updated: Dec 7, 2025
08:05
Detection and Isolation of Cancer in Prostate Biopsies Using Stimulated Raman Histology and Artificial Intelligence
Published on: June 10, 2025
962
Digital Pattern Recognition for the Identification and Classification of Hypospadias Using Artificial Intelligence vs
Nicolas Fernandez1, Armando J Lorenzo2, Mandy Rickard2
1Division of Urology, Seattle Children's Hospital, University of Washington, Seattle, USA.
Urology
|September 29, 2020
Summary
Machine learning and image recognition improve hypospadias classification objectivity. The AI model achieved 90% accuracy, matching expert clinician agreement for classifying distal/proximal hypospadias.
Area of Science:
- Urology
- Medical Imaging
- Artificial Intelligence
Background:
- Hypospadias classification relies on anatomical variables, leading to subjective assessments among clinicians.
- Objective classification is crucial for predicting postoperative outcomes in hypospadias repair.
Purpose of the Study:
- To develop and evaluate a machine learning model for objective hypospadias classification.
- To enhance the objectivity of hypospadias recognition and classification using image recognition.
Main Methods:
- A database of 1169 standardized hypospadias images was used for training a TensorFlow model.
- The model's performance was evaluated on 29 test images and compared against expert clinician classifications.
- Inter- and intrarater reliability analyses were conducted using Fleiss Kappa statistics.
Main Results:
- Model accuracy increased from 60% with 627 images to 90% with 1169 images.
- The image recognition model demonstrated an inter-rater agreement (k=0.86) comparable to expert pediatric urologists (intrarater k=0.74).
Conclusions:
- The developed AI model effectively emulates expert human classification of distal/proximal hypospadias.
- Future applications include standardizing technology use and improving surgical outcome predictions through deep learning algorithms.
