Overview

Schema Version

https://atlas.rsna.org/schemas/2025-11/dataset.json

Name

Two-Stage Training Framework Using Multicontrast MRI Radiomics for IDH Mutation Status Prediction in Glioma

Link

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11294953

Indexing

Keywords: Glioma, Isocitrate dehydrogenase mutation, IDH mutation, Radiomics, MRI

Content: NR, MR, IN, RS

RadLex: RID4026, RID10312

SNOMED: 115240006, 393564001

Author(s)

Nghi C. D. Truong

Chandan Ganesh Bangalore Yogananda

Benjamin C. Wagner

James M. Holcomb

Divya Reddy

Niloufar Saadat

Kimmo J. Hatanpaa

Toral R. Patel

Baowei Fei

Matthew D. Lee

Rajan Jain

Richard J. Bruce

Marco C. Pinho

Ananth J. Madhuranthakam

Joseph A. Maldjian

Organization(s)

The University of Texas Southwestern Medical Center

The University of Texas at Dallas

New York University Grossman School of Medicine

University of Wisconsin–Madison

License

Text: CC BY 4.0

URL: https://creativecommons.org/licenses/by/4.0/

Funding

A.J.M. and J.A.M. supported by NIH/NCI U01CA207091; J.A.M. supported by R01CA260705; R.J.B. supported by NIH 1R01LM013151-01A1.

Ethical review

UTSW Institutional Review Board approval obtained with waiver of consent for retrospective/public data; all internal data anonymized; study HIPAA-compliant.

Date

Published: 2024-05-22

References

[1] Truong NCD, Bangalore Yogananda CG, Wagner BC, et al.. "Two-Stage Training Framework Using Multicontrast MRI Radiomics for IDH Mutation Status Prediction in Glioma". Radiology: Artificial Intelligence. 2024-07-01. doi:10.1148/ryai.230218. PMID: 38775670. PMCID: PMC11294953.

Dataset

Motivation

Develop and evaluate an MRI radiomics framework for preoperative prediction of IDH mutation status in glioma across multi-institutional datasets.

Sampling

Retrospective inclusion from three public and three internal datasets meeting criteria (newly diagnosed glioma, available IDH status, preoperative multicontrast MRI, and tumor segmentation).

Partitioning scheme

Two separate training settings: (1) train on TCIA and test on UTSW, NYU, UWM, UCSF, EGD; (2) train on UTSW and test on TCIA, NYU, UWM, UCSF, EGD. Two-stage training with multiple balanced subsets and model ensembling.

Missing information

Patient-level demographics (age/sex) not reported in the article text; file formats not specified.

Relationships between instances

Multiple MRI sequences per patient; features extracted from two ROIs per patient (whole tumor; NET+NCR+ED).

Noise

Heterogeneity in acquisition protocols, preprocessing pipelines, and tumor mask quality across sources; automated segmentations used without manual correction for several datasets.

External data

Public datasets used: TCGA/IVY (via TCIA), UCSF Preoperative Diffuse Glioma MRI dataset, Erasmus Glioma Database.

Confidentiality

Internal datasets anonymized; HIPAA-compliant.

Re-identification

Data anonymized; no direct identifiers reported.

Sensitive data

Contains medical imaging and molecular status (IDH) from patients with glioma.