Artificial neural networks in urolithiasis

PURPOSE OF REVIEW: The management of urolithiasis is a clinical challenge worldwide which may result in difficulty in diagnosis, treatment and prevention of recurrence. Artificial neural networks (ANNs) are well described adjuncts to many aspects of clinical urological practice. We review literature published in on-line Medline-citable English language journals to assess whether ANNs are useful in clinician-led decision-making processes in urolithiasis. RECENT FINDINGS: Studies have examined the role of ANNs in prediction of stone presence and composition, spontaneous passage, clearance and regrowth after treatment. These reports suggest that ANNs can identify important predictive variables and accurately predict treatment outcome. SUMMARY: Although well described in general urological practice, there is comparatively little research into the role of ANNs in urolithiasis. Initial results appear promising; however, further prospective studies are necessary to determine if this mode of analysis is superior to standard statistical predictive methods.