Ratio maps of T1w/T2w MRI signal intensity do not improve deep-learning segmentation of pediatric brain tumors
Overall, we demonstrate that T1w/T2w ratio maps do not improve deep learning models for segmenting pediatric brain tumor subregions using nnU-Net, despite their strong biophysical basis for tissue discrimination. T hese findings suggest that such data augmentation strategies may not provide added value and highlight the importance of rigorous validation in medical imaging research.
Authors
- Heather Rose (ORCID: https://orcid.org/0000-0002-0346-1334)
- Daniel Griffiths-King (ORCID: https://orcid.org/0000-0001-5797-9203)
- Timothy Mulvany (ORCID: https://orcid.org/0009-0000-1699-4415)
- Jan Novak
Institutions
- Nottingham University Hospitals NHS Trust (GB)
- Aston University (GB)
- University of Birmingham (GB)
Publication Details
- Journal
- PLoS ONE
- Published
- 2025-12-22
- DOI
- https://doi.org/10.1371/journal.pone.0323398
- Primary Topic
- Glioma Diagnosis and Treatment
- Type
- article
- Field-Weighted Citation Impact
- 0.00