Related Experiment Video
Updated: Jan 6, 2026
03:13
Author Spotlight: A Single-Entry Point Endoscopic Intraventricular Approach for Third Ventriculostomy and Pineal Biopsy
Published on: June 28, 2024
1.2K
Machine learning for endoscopic third ventriculostomy success prediction-a systematic review and meta-analysis
Anna Łajczak1, Yasmin Picanço Silva2, Paweł Łajczak3
1Department of Biophysics, Faculty of Medical Sciences in Zabrze, Medical University of Silesia in Katowice, Katowice, Poland.
Summary
Machine learning (ML) models show moderate potential for predicting endoscopic third ventriculostomy (ETV) success in pediatric hydrocephalus. However, they do not currently outperform traditional ETVSS and logistic regression models.
Area of Science:
- Neurosurgery
- Medical Informatics
- Pediatric Neurology
Background:
- Predicting endoscopic third ventriculostomy (ETV) success for pediatric obstructive hydrocephalus is challenging.
- Traditional tools like ETVSS and logistic regression (LR) are widely used but have limitations.
- Machine learning (ML) offers potential for improved prediction accuracy.
Purpose of the Study:
- To systematically review and meta-analyze the effectiveness of ML models in predicting ETV success.
- To compare the performance of ML models against traditional predictive tools.
Main Methods:
- Systematic search of five databases for studies using ML algorithms to predict ETV success.
- Inclusion of studies reporting the area under the receiver operating characteristic curve (AUC).
- Definition of ETV success as absence of failure criteria within 6 months (recurrence, re-operation, mortality).
Main Results:
- Four studies with 3087 pediatric patients were included.
- Pooled AUC for ML models was 0.63 (95% CI 0.56-0.70) with high heterogeneity (I²=96%).
- Models incorporating imaging data showed a higher AUC (0.74), but no significant difference was found compared to traditional tools.
Conclusions:
- ML models demonstrate moderate potential for predicting ETV success.
- Current ML models do not outperform traditional ETVSS and LR models in clinical practice.
- Further research is needed due to high heterogeneity and methodological limitations.
