Abstract
Automated segmentation of pediatric high-grade brain tumors
on multiparametric magnetic resonance imaging (MRI) remains
challenging because tumor appearance varies across diagnoses, institutions,
and treatment stages. This paper presents the design and official
results of the 2026 Multi-Consortium International Pediatric Brain Tumor
Segmentation (BraTS-PEDs) Challenge. The retrospective, multi-institutional
dataset includes treatment-naive and post-treatment imaging,
with expert-reviewed reference standard annotations for six tumor
regions. The cohort was divided into 294 training cases with released annotations,
91 validation cases with withheld annotations, and 109 hidden
test cases. Seventeen containerized algorithms submitted by participants
were evaluated using the Dice similarity coefficient (DSC) and normalized
surface distance (NSD) across six tumor regions. Final rankings were
determined using PermRanker, which aggregates case-wise metric ranks
and performs pairwise permutation testing for statistical significance.
The team ranked first achieved the highest mean DSC (0.930±0.112)
and NSD (0.849±0.197) on the whole tumor. These results establish
a standardized benchmark for developing and comparing clinically relevant
pediatric brain tumor segmentation methods. Future work will
provide deeper insights into these methods and assess the potential of
AI-based segmentation for pediatric brain tumors through an extensive
post-challenge analysis.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/BraTS_PEDs_000.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: Not Available
BibTex
@InProceedings{JiaZhi_MultiConsortium_MICCAISAT2026,
author = { Jiang, Zhifan AND Fathi Kazerooni, Anahita AND Astaraki, Mehdi AND Bakas, Spyridon AND Chung, Verena AND Farahani, Keyvan AND Correia de Verdier, Maria AND Baid, Ujjwal AND LaBella, Dominic AND Conte, Gian Marco AND You, Suhang AND Yordanov, Nikolay AND Aboian, Mariam S. AND Buzduga, Paula AND Grosskopf, Erik AND Kofler, Florian AND Wiestler, Benedikt AND Menze, Bjoern AND Huang, Raymond Y. AND Batal, Fadel AND Varshochi, Sanaz AND Adib Moradi, Sahand AND Sheth, Nakul AND Nandolia, Khanak AND Sánchez-Montaño, Mariana AND Ware, Jeffrey B. AND Nabavizadeh, Ali AND Vossough, Arastoo AND Linguraru, Marius George},
title = { { Multi-Consortium International Pediatric Brain Tumor Segmentation Challenge (BraTS-PEDs) 2026: Design and Results } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17253},
month = {pending},
page = {pending}
}
