A team of investigators from the UC San Francisco (UCSF) and the University of Pennsylvania aimed to expand and improve upon the available datasets in the hopes of continuing to push the boundaries of AI applications for diffuse gliomas. The resulting data resource, titled " The University of California San Francisco Preoperative Diffuse Glioma MRI Dataset," is discussed in Radiology: Artificial Intelligence.
The University of California San Francisco Preoperative Diffuse Glioma MRI (UCSF-PDGM) dataset includes 500 subjects with grade 2-4 diffuse gliomas and includes standardized 3-T three-dimensional preoperative MRI protocol, diffusion MRI, and perfusion MRI, multicompartment tumor segmentations, tumor genetic data and treatment and survival data.
"The UCSF-PDGM not only significantly increases the total number of publicly available diffuse glioma MRI cases, but also provides a unique contribution in terms of MRI technique," write the investigators, led by Evan Calabrese, MD, PhD, a former resident and fellow at UCSF Radiology and current assistant professor of Radiology at Duke University Medical Center. "The inclusion of 3D sequences and advanced MRI techniques like ASL and HARDI provides a new opportunity for researchers to explore the potential utility of cutting-edge imaging for AI applications."
The team was compiled of investigators from the UCSF Center for Intelligent Imaging (UCSF ci2), the UCSF Department of Radiology and Biomedical Imaging and the Center for Biomedical Image Computing and Analytics (CBICA) at the University of Pennsylvania. They assert that the UCSF-PDGM dataset, when combined with other publicly available datasets, could have an impact on the next phase of radiologic AI research on diffuse gliomas.
"However, the UCSF-PDGM dataset's potential will only be realized if the radiology AI research community takes advantage of this new data resource," write the investigators. "We hope that this dataset sparks inspiration in the next generation of AI researchers, and we look forward to the new techniques and discoveries that the UCSF-PDGM will generate."
The investigators' research was partly supported by awards from the National Institutes of Health and the Radiological Society of North American Research and Education Foundation.
Additional authors from UCSF ci2 include senior author Christopher Hess, MD, PhD, the founding director of UCSF ci2 and the chair of UCSF Radiology, Javier Villanueva-Meyer, MD, Jeffrey Rudie, MD, PhD, Andreas Rauschecker, MD, PhD, and John Mongan, MD, PhD. UCSF Radiology's Soonmee Cha, MD, and CBICA's Ujjwal Baid, PhD, and Spyridon Bakas, PhD, also contributed.
The UCSF-PDGM data is publicly available on The Cancer Imaging Archive.
