Systematic Review and Meta-Analysis of AI Prediction Models for Prostate Cancer

A systematic review and meta-analysis of 144 studies evaluating artificial intelligence models for predicting prostate cancer outcomes. The meta-analysis includes 85 studies and reports pooled area under the curve (AUC) metrics for endpoints including overall survival, progression-free survival, and recurrence distance metastasis. The review was conducted by Emmanuel Nsedu Israel and registered in PROSPERO (CRD42025611480).

Use Cases

• Benchmarking new prostate cancer prediction models based on reported AUC metrics for endpoints like overall survival.
• Assessing methodological quality of clinical AI studies based on the Prediction Model Quality Score (PMQS) framework.
• Identifying performance gaps in AI models for specific prediction endpoints such as toxicity or quality of life.
• Informing systematic review protocols for clinical prediction models based on the PRISMA criteria used.

Strengths

• Includes a meta-analysis of 85 studies with pooled AUC results for five distinct prediction endpoints.
• Methodological quality of included studies was assessed using the Prediction Model Quality Score (PMQS).
• Publication bias was assessed using funnel plots, Egger's regression test, and the trim-and-fill method.

Limitations

• The dataset is a 18.2 KB DOCX file, which is a small document rather than a structured data table.
• Row and column-level data are unavailable; the file contains a review summary, not raw model performance data.
• The description does not specify the temporal or geographic scope of the underlying studies reviewed.

Provenance

Source

figshare

Collection Method

Systematic review and meta-analysis of studies from PubMed and Scopus.

Freshness

Last updated 2026-05-01 05:28:16.