Computer aided diagnosis in acute testicular pain

Area of Science:

• Urology
• Medical Informatics

Background:

• Testicular torsion and epididymo-orchitis present similar symptoms, complicating diagnosis.
• Delayed diagnosis of testicular torsion can lead to testicular loss.

Purpose of the Study:

• To develop and evaluate a computer program for differentiating testicular torsion from epididymitis.
• To assess the program's impact on diagnostic accuracy and surgical exploration rates.

Main Methods:

• A computer program was developed using a literature-derived database.
• The program was retrospectively analyzed on 495 patients with suspected testicular pathologies.
• Diagnostic accuracy was compared against final diagnoses and junior/senior clinician performance.

Main Results:

• The computer program achieved 92.3% accuracy in differentiating torsion from epididymitis.
• Implementing the program could reduce epididymitis exploration rates from 18% to 1.8%.
• The program enhanced junior doctor diagnostic accuracy to senior colleague levels, with similar missed torsion rates.

Conclusions:

• A computer-assisted diagnostic tool can significantly improve accuracy in acute testicular pain management.
• This non-invasive, inexpensive technique offers potential to increase testicular salvage rates.
• The program aids in reducing unnecessary surgical interventions for epididymitis.