Overview

Schema Version

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

Name

University of California San Francisco Preoperative Diffuse Glioma MRI (UCSF-PDGM) dataset

Link

https://doi.org/10.7937/tcia.bdgf-8v37

Indexing

Keywords: Informatics, MR Diffusion Tensor Imaging, MR Perfusion, MR Imaging, Neuro-Oncology, CNS, Brain/Brain Stem, Oncology, Radiogenomics, Radiology-Pathology Integration

Content: NR, MR, OI

RadLex: RID10312, RID6677, RID38778

SNOMED: 1163375002

Author(s)

Evan Calabrese

Javier E. Villanueva-Meyer

Jeffrey D. Rudie

Andreas M. Rauschecker

Ujjwal Baid

Spyridon Bakas

Soonmee Cha

John T. Mongan

Christopher P. Hess

Organization(s)

Center for Intelligent Imaging (Ci2), Department of Radiology & Biomedical Imaging, University of California San Francisco

Center for Biomedical Image Computing and Analytics (CBICA), University of Pennsylvania

Contact

Corresponding author: Evan Calabrese (evan.calabrese@ucsf.edu)

Funding

Supported in part by NIH awards NCI:U01CA242871 and T32EB001631, and by the RSNA Research & Education Foundation award RR2011.

Ethical review

Data collection approved by the UCSF institutional review board with waiver of consent.

Comments

Publicly available preoperative diffuse glioma MRI dataset from a single center with standardized 3-T protocol, advanced MRI (ASL, HARDI), multicompartment tumor segmentations, genetic biomarkers, and treatment/survival metadata.

Date

Published: 2022-10-05

References

[1] Calabrese E, Villanueva-Meyer JE, Rudie JD, Rauschecker AM, Baid U, Bakas S, Cha S, Mongan JT, Hess CP. "The University of California San Francisco Preoperative Diffuse Glioma MRI Dataset". Radiology: Artificial Intelligence. 2022-11-01. doi:10.1148/ryai.220058. PMID: 36523646. PMCID: PMC9748624.

Dataset

Motivation

To provide a large, standardized, multi-sequence 3-T preoperative diffuse glioma MRI dataset with advanced imaging, tumor segmentations, genetics, and outcomes to foster AI research.

Sampling

Single-center cohort of 501 adult patients with histopathologically confirmed WHO grade 2–4 diffuse gliomas imaged preoperatively between 2015 and 2021; excluded prior treatment other than biopsy (allowed in 69/501).

Relationships between instances

Each patient has multiple coregistered 3D MRI sequences, derived diffusion maps, and multicompartment tumor segmentations linked to genetic biomarkers and clinical outcomes.

Noise

Some cases excluded for severe artifacts; included cases underwent quality review (including BraTS 2021 review for 387/501).

External data

A portion of the dataset is included in the 2021 BraTS challenge dataset.

Confidentiality

Public dataset available via The Cancer Imaging Archive (TCIA).

Sensitive data

Includes genetic biomarker status (IDH, MGMT for grade 3–4, 1p/19q for most) and survival/treatment details.