Machine Learning for Predicting Ki-67 Expression in Renal Tumors: A Systematic Review And Meta-Analysis.

In 30 seconds

This systematic review and meta-analysis evaluated the diagnostic performance of machine learning models in predicting Ki-67 expression in renal tumors, analyzing data from seven retrospective cohort studies involving 1176 training and 885 validation patients. The pooled sensitivity and specificity for validation cohorts were 0.83 and 0.73, respectively, indicating moderate diagnostic performance.

Key findings

• Pooled sensitivity for training cohorts was 0.81 and for validation cohorts was 0.83.
• Pooled specificity for training cohorts was 0.84 and for validation cohorts was 0.73.
• The area under the curve (AUC) was 0.85 for training cohorts and 0.86 for validation cohorts.
• Specificity showed substantial heterogeneity with I² > 74%.

Why it matters

Understanding the diagnostic performance of machine learning models for Ki-67 expression can aid clinicians in evaluating noninvasive tools that may complement traditional pathological assessments, potentially improving patient management in renal tumors.