MétaCan
Menu
← Back to cohort

Exploring the Impact of Supervised Multimodal Learning on the Performance and Explainability of Pediatric Brain Tumor Molecular Diagnosis

2025· article· W7110140606 on OpenAlexaff

Bibliographic record

Venuenot available
Typearticle
Language
FieldMedicine
TopicGlioma Diagnosis and Treatment
Canadian institutionsHospital for Sick Children
Fundersnot available
KeywordsConvolutional neural networkSegmentationMagnetic resonance imagingDeep learningMedical imagingNeuroimagingBrain tumorRadiogenomicsArtificial neural network

Abstract

fetched live from OpenAlex

Pediatric Low-grade Glioma (pLGG) is one of the most common brain tumors in children, and identifying its genetic markers (molecular diagnosis) is crucial for tumor prognosis and targeted treatment planning. Convolution Neural Networks (CNNs) have shown strong performance in predicting these genetic markers from brain magnetic resonance imaging (MRI) data. Nonetheless, most CNN-based architectures rely on tumor segmentation masks to specify regions of interest (ROIs) in the image to achieve high performance, which are labor-intensive and costly to obtain. In this work, we propose a supervised multimodal learning framework that integrates radiology reports, readily available in most medical imaging datasets, with entire MRI scans, eliminating the need for providing segmentation masks. In addition to improving the performance of pLGG molecular diagnosis, we examine the effect of this framework on the alignment between predictive imaging features and domain knowledge, as a measure of model explainability. Our results indicate a considerable improvement in the Area Under the Receiver Operating Curve compared to an MRI-only CNN model (0.863 vs 0.79), highlighting the value of radiology reports in enhancing CNN performance. However, the lower dice score between model attention maps and segmentation masks suggests that supervised training may not optimally integrate the two modalities. These findings signify the potential of multimodal learning in pLGG molecular diagnosis while underscoring the need for future exploration of other approaches, e.g., self-supervised learning, to improve the alignment between learnt features and clinical reasoning.

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 imitation

Not 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.

metaresearch head score (Codex)0.008
metaresearch head score (Gemma)0.025
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.008
Threshold uncertainty score0.041

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0080.025
Meta-epidemiology (narrow)0.0020.000
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0010.001
Science and technology studies0.0000.001
Scholarly communication0.0010.002
Open science0.0010.001
Research integrity0.0010.002
Insufficient payload (model declined to judge)0.0010.000

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.

Opus teacher head0.033
GPT teacher head0.285
Teacher spread0.251 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designSimulation or modeling
Domainnot available
GenreEmpirical

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".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same topicGlioma Diagnosis and Treatment→French-language works237,207→