[PDF] The University of Pennsylvania glioblastoma (UPenn-GBM) cohort: advanced MRI, clinical, genomics, & radiomics | Semantic Scholar (2024)

A single-center longitudinal GBM MRI dataset with expert ratings of selected follow-up studies according to the response assessment in neuro-oncology criteria (RANO) is released with details about the rationale of the ratings.

This study takes a third-party database and reduces physician bias from interfering with study findings, and evaluates the neuroradiological parameters of de novo GBM by analyzing the brain multi-parametric magnetic resonance imaging scans acquired from a publicly available database analysis of the scans.

Unsupervised joint machine learning between radiomic and genomic data is developed, thereby identifying distinct glioblastoma subtypes, demonstrating the synergistic value of integrated radiomic signatures and molecular characteristics for glioblastoma subtyping.

Compared to radiomics and shallow machine learning, deep learning can be a more robust, non-invasive, and effective approach, providing more valuable information as clinicians develop personalized medical protocols for glioblastoma patients.

A novel pipeline, Brain Radiology Aided by Intelligent Neural NETworks (BRAINNET), which leverages MaskFormer, a vision transformer model, and generates robust tumor segmentation maks is proposed, which achieves state-of-the-art results in segmenting all three different tumor regions.

The proposed method demonstrates that a fully automatic mpMRI analysis using deep learning can successfully predict tumor infiltration in peritumoral region for GBM patients while bypassing the intensive requirement for expert-drawn ROIs.

The proposed transformer-based survival prediction model integrates complementary information from diverse input modalities, contributing to improved glioblastoma survival prediction compared to state-of-the-art methods.

This study critically addresses the challenges associated with locating, preprocessing, and extracting quantitative features from open-access datasets, to facilitate more robust and reliable evaluations of radiomics models.

This challenge will provide a community standard and benchmark for state-of-the-art automated intracranial meningioma segmentation models based on the largest expert annotated multilabel meninguoma mpMRI dataset to date.

...

...

In multivariable survival models, rCBVNER provided unique prognostic information that went above and beyond the assessment of all NER imaging features, as well as clinical and genomic features.

Results indicate that quantitative multivariate analysis of features extracted from clinically-acquired MRI may provide a radiographic biomarker of the transcriptomic profile of glioblastoma.

Imaging signatures of presurgical MP-MRI scans reveal relatively high predictability of time and location of GBM recurrence, subject to the patients receiving standard first-line chemoradiation therapy.

This quantitative evaluation indicates that accurate survival prediction in glioblastoma patients is feasible using solely Bas-mp MRI and integrative advanced radiomic features, which can compensate for the lack of Adv-mpMRI.

This work addresses two critical challenges with regard to developing robust radiomic approaches: the lack of availability of reliable segmentation labels for GBM tumor sub-compartments, and identifying "reproducible" radiomic features that are robust to segmentation variability across readers/sites.

This set of labels and features should enable direct utilization of the TCGA/TCIA glioma collections towards repeatable, reproducible and comparative quantitative studies leading to new predictive, prognostic, and diagnostic assessments, as well as performance evaluation of computer-aided segmentation methods, and comparison to the state-of-the-art method.

This analysis demonstrates a method for consistent image feature annotation capable of reproducibly characterizing brain tumors and shows that radiologists' estimations of macroscopic imaging features can be combined with genetic alterations and gene expression subtypes to provide deeper insight to the underlying biologic properties of GBM subsets.

It is hypothesized that integrative analysis of multi-parametric magnetic resonance imaging (mpMRI) via multivariate machine learning (ML), will enhance subtle yet important radiographic HGG characteristics, and reveal imaging signatures determinant of IDH1 mutational status.

...

...