Related Experiment Video
Updated: Jul 9, 2025
05:37
An R-Based Landscape Validation of a Competing Risk Model
Published on: September 16, 2022
2.1K
Development and external validation of clinical prediction models for pituitary surgery
Olivier Zanier1, Matteo Zoli2,3, Victor E Staartjes1
1Machine Intelligence in Clinical Neuroscience (MICN) Laboratory, Department of Neurosurgery, Clinical Neuroscience Center, University Hospital Zurich, University of Zurich, Zurich, Switzerland.
Brain & Spine
|November 29, 2023
Summary
Machine learning models accurately predict surgical outcomes for pituitary adenoma (PA) patients. These models, using preoperative data, aid in personalized surgical strategies for gross total resection (GTR), biochemical remission (BR), and improved pituitary function (IMP).
Area of Science:
- Neurosurgery
- Endocrinology
- Medical Artificial Intelligence
Background:
- Pituitary adenoma (PA) resection aims for gross total resection (GTR), biochemical remission (BR), and improved pituitary function (IMP).
- Predicting these outcomes preoperatively can enhance patient care and personalize treatment strategies.
Purpose of the Study:
- To develop and validate machine learning (ML) models for predicting GTR, BR, and IMP in PA surgery.
- To utilize simple, preoperatively available data for predictive modeling.
Main Methods:
- ML models were developed using data from 1203 patients undergoing endoscopic transsphenoidal surgery for PAs in Bologna, Italy.
- External validation was performed on a separate cohort of 206 patients from Zurich, Switzerland.
Main Results:
- Models demonstrated good generalizability with AUCs at external validation of 0.72 for GTR, 0.69 for BR, and 0.82 for IMP.
- The study achieved GTR in 58.7% of the external validation cohort, BR in 22.3%, and IMP in 20.4%.
Conclusions:
- ML models show potential for predicting key outcomes in pituitary adenoma surgery.
- These predictive capabilities can support tailored surgical approaches for individual patients.
