Related Experiment Video
Updated: Dec 15, 2025
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
Published on: December 15, 2023
Automatic Machine Learning to Differentiate Pediatric Posterior Fossa Tumors on Routine MR Imaging
1Department of Neurology (H.Z., L.T., B.X.), Xiangya Hospital of Central South University, Changsha, Hunan, China.
Background And Purpose:
Differentiating the types of pediatric posterior fossa tumors on routine imaging may help in preoperative evaluation and guide surgical resection planning. However, qualitative radiologic MR imaging review has limited performance. This study aimed to compare different machine learning approaches to classify pediatric posterior fossa tumors on routine MR imaging.
Materials And Methods:
This retrospective study included preoperative MR imaging of 288 patients with pediatric posterior fossa tumors, including medulloblastoma (n = 111), ependymoma (n = 70), and pilocytic astrocytoma (n = 107). Radiomics features were extracted from T2-weighted images, contrast-enhanced T1-weighted images, and ADC maps. Models generated by standard manual optimization by a machine learning expert were compared with automatic machine learning via the Tree-Based Pipeline Optimization Tool for performance evaluation.
Results:
For 3-way classification, the radiomics model by automatic machine learning with the Tree-Based Pipeline Optimization Tool achieved a test micro-averaged area under the curve of 0.91 with an accuracy of 0.83, while the most optimized model based on the feature-selection method χ2 score and the Generalized Linear Model classifier achieved a test micro-averaged area under the curve of 0.92 with an accuracy of 0.74. Tree-Based Pipeline Optimization Tool models achieved significantly higher accuracy than average qualitative expert MR imaging review (0.83 versus 0.54, P < .001). For binary classification, Tree-Based Pipeline Optimization Tool models achieved an area under the curve of 0.94 with an accuracy of 0.85 for medulloblastoma versus nonmedulloblastoma, an area under the curve of 0.84 with an accuracy of 0.80 for ependymoma versus nonependymoma, and an area under the curve of 0.94 with an accuracy of 0.88 for pilocytic astrocytoma versus non-pilocytic astrocytoma.
Conclusions:
Automatic machine learning based on routine MR imaging classified pediatric posterior fossa tumors with high accuracy compared with manual expert pipeline optimization and qualitative expert MR imaging review.
Insights
Machine learning accurately classifies pediatric posterior fossa tumors using routine MRI scans. Automatic machine learning models outperformed expert reviews, improving diagnostic accuracy for these critical brain tumors.
Area of Science:
- Neuro-oncology
- Radiology
- Artificial Intelligence
Background:
- Differentiating pediatric posterior fossa tumors is crucial for surgical planning.
- Qualitative MRI review has limitations in diagnostic performance.
- Machine learning offers potential for improved classification accuracy.
Purpose of the Study:
- To compare machine learning approaches for classifying pediatric posterior fossa tumors.
- To evaluate the performance of automatic vs. manual machine learning optimization.
- To assess the utility of radiomics features from routine MR imaging.
Main Methods:
- Retrospective analysis of 288 pediatric posterior fossa tumor MRIs.
- Extraction of radiomics features from T2-weighted, contrast-enhanced T1-weighted images, and ADC maps.
- Comparison of standard manual machine learning optimization with automatic machine learning (Tree-Based Pipeline Optimization Tool).
Main Results:
- Automatic machine learning achieved high accuracy (0.83) and AUC (0.91) for 3-way classification.
- Automatic machine learning significantly outperformed qualitative expert review (0.83 vs. 0.54, P < .001).
- High AUCs and accuracies were observed for binary classifications of specific tumor types.
Conclusions:
- Automatic machine learning on routine MR imaging effectively classifies pediatric posterior fossa tumors.
- This approach demonstrates superior accuracy compared to manual optimization and expert review.
- Machine learning holds promise for enhancing preoperative evaluation and surgical guidance.
