Reducing Gadolinium Dependency in Neurooncology: A Multicenter Machine-Learning Driven Estimation of Brain Glioma Enhancement from Non-Contrast T1 MRI
Bibliographic record
Abstract
Gliomas, including low-grade gliomas (LGG) and glioblastoma (GBM), require precise imaging. Gadolinium-based contrast-enhanced (GBCA) MRI is routinely used but poses risks such as nephrogenic systemic fibrosis and gadolinium retention, particularly in vulnerable patients. This study investigates a Radiomics-based (RF) machine learning (ML) framework to estimate tumor contrast enhancement likelihood from non-contrast T1-weighted MRI, eliminating gadolinium use while maintaining diagnostic accuracy. A 5-fold cross-validation training process was performed on multi-institutional datasets (231 samples from Brats_Africa, 59 from TCGA_LGG, and 496 from UCSF_PDGM). External nested testing was performed on a separate dataset from a different center, consisting of 660 samples from the UPENN-GBM cohort. Our pipeline evaluated 63 ML models, including 7 feature selection algorithms such as Mutual Information (MIS), Embedded Feature Importance via Extra Trees Classifier (EFIET), and others, linked with 9 classifiers, including Random Forest (RandF), Multilayer Perceptron (MLP), and others. Images were normalized, and tumor masks were verified by two medical professionals. An Experienced radiologist labeled MRI images using non-contrast and contrast-enhanced T1-weighted scans, classifying enhanced as 1 and non-enhanced as 0. A total of 107 RFs were extracted from non-contrast T1-weighted MRI using PyRadiomics. The top-performing predictive model (EFIET feature selector + MLP classifier) achieved an average accuracy of$0.94 \pm 0.02$, with an external nested test of$0.87 \pm 0.01$. Overall, the 54 RFs selected by EFIET, when applied to different classifiers, outperformed other subsets. This study offers a safer, cost-effective alternative to GBCA MRI, reducing scan time and avoiding contrast-related risks. Our framework highlights the transformative potential of ML and radiomics in neuro-oncology, particularly for vulnerable populations and resource-limited settings.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".