Multi-Institutional Annotated Multiparametric MRI Dataset of Pediatric High-Grade Gliomas
Radiology: Artificial Intelligence
Anahita Fathi Kazerooni, PhD, Zhifan Jiang, PhD, Deep Gandhi, MSc, Nastaran Khalili, MD, Xinyang Liu, PhD, Wenxin Tu, BSc, Jeffrey B. Ware, MD, Ariana M. Familiar, PhD, Bhavyasri Vunnava, MD, Anna Zapaishchykova, PhD, Aaron S. McAllister, MD, Mariana Sanchez-Montano, MD, MSc, Nakul Sheth, MD, Khanak K. Nandolia, MD, Hollie Anne Lai, MD, Julija Pavaine, MD, Sanjay P. Prabhu, MD, Debanjan Haldar, MD, MSc, Sanaz Varshochi, MD, Sina Bagheri, MD, Hannah Anderson, MSc, Shuvanjan Haldar, BSc, Neda Khalili, MD, Anurag Gottipati, BSc, Ibraheem Salman Shaikh, MD, Ethan Castellino, MD, Avani Mangoli, MD, Harsh Gohil, MD, Nazanin Maleki, MD, Justin Low, MD, Trent Hummel, MD, Roger J. Packer, MD, Andrea Franson, MD, Phillip B. Storm, MD, Spyridon Bakas, PhD, Evan Calabrese, MD, Mariam Aboian, MD, Peter de Blank, MD, Benjamin H. Kann, MD, Brian Rood, MD, Adam C. Resnick, PhD, Ali Nabavizadeh, MD, Arastoo Vossough, MD, and Marius George Linguraru, PhD
Summary
The BraTS-PEDs dataset provides the largest, multi-institutional, publicly available collection of annotated multiparametric MRI scans of pediatric high-grade gliomas, enabling standardized benchmarking and development of AI tools in pediatric neuro-oncology.
Key Points
- The Brain Tumor Segmentation in Pediatrics (BraTS-PEDs) dataset, shared via The Cancer Imaging Archive, includes multiparametric MRI and expert-refined tumor subregion annotations from 457 pediatric patients with high-grade gliomas collected across multiple international institutions.
- A standardized preprocessing pipeline and multistage expert annotation process were applied, providing harmonized imaging data and reproducible reference standard segmentations aligned with Response Assessment in Pediatric Neuro-Oncology recommendations.
- The publicly available dataset enables benchmarking, development, validation, and generalization testing of artificial intelligence models for pediatric brain tumor segmentation across heterogeneous, multi-institutional imaging data.
Citation
Kazerooni, Anahita Fathi, et al. “Multi-Institutional Annotated Multiparametric MRI Dataset of Pediatric High-Grade Gliomas.” Radiology: Artificial Intelligence 8.3 (2026): e250902.