A systematic review of the applications of Expert Systems (ES) and machine learning (ML) in clinical urology

Abstract Background Testing a hypothesis for ‘factors-outcome effect’ is common quest, but standard statistical regression analysis tools are rendered ineffective by data contaminated with too many noisy variables. Expert Systems (ES) can provide an alternative methodology in analysing to identify variables the highest correlation outcome. By applying their effective machine learning (ML) abilities, significant research time and costs be saved. The study aims systematically review applications of ES urological methodological models multi-variate analysis. Their domains, development validity will identified. Methods PRISMA was applied formulate method gathering This search included seven most relevant information sources: WEB OF SCIENCE, EMBASE, BIOSIS CITATION INDEX, SCOPUS, PUBMED, Google Scholar MEDLINE. Eligible articles were if they one known ML clear question involving multivariate Only pertinent methods included. analysed system model, applications, input/output variables, target user, validation, outcomes. Both variable comparatively reported each system. Results identified n = 1087 from all databases 712 eligible examination against inclusion criteria. A total 168 systems finally demonstrating recent increase uptake academic urology particular artificial neural networks 31 systems. Most oncology (prostate cancer 15, bladder 13) where diagnostic, prognostic survival predictor markers investigated. Due heterogeneity tests, meta-analysis not feasible. Conclusion utility offers potential have demonstrated valid model complexity challenge clinics whilst limitation this domain has created gap further studies. Integration computer scientists units promoted use clinical research.