Related Experiment Video
Updated: Mar 28, 2026
Establishment of Orthotopic Patient-derived Xenograft Models for Brain Tumors using a Stereotaxic Device
Published on: May 2, 2025
Enhancing efficiency in pediatric brain tumor segmentation using a pathologically diverse single-center clinical
Arianna Piffer1, Josef Alois Buchner2, Antonio Giulio Gennari3,4
1Division of Oncology and Children's Research Center, University Children's Hospital Zurich, Zurich, Switzerland.
Insights
Deep learning segmentation accurately delineates whole pediatric brain tumors (PBTs) and T2-hyperintensity, comparable to human experts. Further refinement is needed for enhancing tumor and cystic component segmentation in PBTs.
Area of Science:
- Medical imaging analysis
- Artificial intelligence in oncology
Background:
- Pediatric brain tumors (PBTs) are diverse and challenging to diagnose and treat.
- Deep learning (DL) shows promise for PBT segmentation but requires validation across subtypes and MRI protocols.
Purpose of the Study:
- To evaluate the performance of a 3D nnU-Net model for segmenting various pediatric brain tumor subregions.
- To assess DL segmentation accuracy across different PBT subtypes, MRI sequences, and tumor locations.
Main Methods:
- Retrospective analysis of 174 pediatric patients with diverse brain tumors.
- Utilized MRI sequences (T1, T1-C, T2, FLAIR) with manual annotations for whole tumor (WT), T2-hyperintensity (T2H), enhancing tumor (ET), and cystic component (CC).
- Trained and tested a 3D nnU-Net model, evaluating performance using Dice Similarity Coefficient (DSC) against human variability.
Main Results:
- The DL model achieved robust performance for WT and T2H segmentation (mean DSC: 0.85), comparable to human annotators.
- Moderate accuracy was observed for ET segmentation (mean DSC: 0.75), with poor performance for CC (mean DSC: 0.26).
- Segmentation accuracy varied by tumor type, MRI sequence combination, and location; T1, T1-C, and T2 sequences yielded results similar to the full protocol.
Conclusions:
- Deep learning-based segmentation is feasible and effective for WT and T2H in PBTs.
- Challenges persist in segmenting ET and CC, necessitating further model development.
- Findings suggest potential for protocol simplification and automated volumetric assessment in pediatric neuro-oncology.
Background:
Brain tumors are the most common solid malignancies in children, encompassing diverse histological, molecular subtypes, and imaging features and outcomes. Pediatric brain tumors (PBTs), including high- and low-grade gliomas (HGG and LGG), medulloblastomas (MB), ependymomas, and rarer forms, pose diagnostic and therapeutic challenges. Deep learning (DL)-based segmentation offers promising tools for tumor delineation, yet its performance across heterogeneous PBT subtypes and MRI protocols remains uncertain.
Methods:
A retrospective single-center cohort of 174 pediatric patients with HGG, LGG, MB, ependymomas, and other rarer subtypes was used. MRI sequences included T1, T1 post-contrast (T1-C), T2, and FLAIR. Manual annotations were provided for 4 tumor subregions: whole tumor (WT), T2-hyperintensity (T2H), enhancing tumor (ET), and cystic component (CC). A 3D nnU-Net model was trained and tested (121/53 split), with segmentation performance assessed using the Dice similarity coefficient (DSC) and compared against intra- and interrater variability.
Results:
The model achieved robust performance for WT and T2H (mean DSC: 0.85), comparable to human annotator variability (mean DSC: 0.86). ET segmentation was moderately accurate (mean DSC: 0.75), while CC performance was poor (mean DSC: 0.26). Segmentation accuracy varied by tumor type, MRI sequence combination, and location. Notably, T1, T1-C, and T2 combined produced results nearly equivalent to the full protocol.
Conclusions:
DL-based segmentation is feasible for PBTs, particularly for T2H and WT. Challenges remain for ET and CC segmentation, highlighting the need for further refinement. These findings support the potential for protocol simplification and automation to enhance volumetric assessment and streamline pediatric neuro-oncology workflows.
