analiza aiqbench
NeoTriFuse introduces a novel framework for neonatal mortality prediction that treats data missingness not as noise to be cleaned, but as an explicit reliability signal to guide multimodal fusion. By dynamically modulating contributions from static variables and temporal encoders based on observation completeness, the model achieves robust performance despite extreme class imbalance and heterogeneous clinical factors. This approach advances the practical deployment of machine learning in critical care settings where data integrity is inherently variable.
NeoTriFuse: Reliability-Aware Multimodal Fusion under Missingness Heterogeneity for Neonatal Mortality Risk Prediction
arXiv:2608.26436v1 Announce Type: new Abstract: Neonatal mortality risk prediction from bedside monitoring data remains challenging due to extreme class imbalance, heterogeneous clinical risk factors, multi-scale temporal dynamics, and substantial missingness. We propose NeoTriFuse, a reliability-aware multimodal fusion framework for missingness-heterogeneous neonatal monitoring data. Unlike conventional multimodal approaches that treat missingness primarily as a preprocessing issue, NeoTriFuse models missingness as an explicit reliability signal that dynamically modulates modality contributions during fusion. The framework integrates static perinatal variables, local-global temporal encoders, and patient-level statistical summaries through reliability-guided gating mechanisms, while jointly optimizing mortality prediction and an auxiliary length-of-stay objective. NeoTriFuse achieves competitive performance, with an F1 score of 0.6736 +/- 0.0216 and an AUROC of 0.9454 +/- 0.0056. Ablation studies indicate that the local-global temporal architecture and patient-level summary branch contribute most substantially to predictive performance, while reliability-aware gating provides additional improvements on threshold-dependent metrics under heterogeneous observation completeness. Sensitivity analyses further suggest stable performance across nearby hyperparameter settings. Overall, the findings support reliability-aware multimodal fusion as a practical approach for neonatal mortality prediction under realistic clinical missingness conditions.
Abstract:Neonatal mortality risk prediction from bedside monitoring data remains challenging due to extreme class imbalance, heterogeneous clinical risk factors, multi-scale temporal dynamics, and substantial missingness. We propose NeoTriFuse, a reliability-aware multimodal fusion framework for missingness-heterogeneous neonatal monitoring data. Unlike conventional multimodal approaches that treat missingness primarily as a preprocessing issue, NeoTriFuse models missingness as an explicit reliability signal that dynamically modulates modality contributions during fusion. The framework integrates static perinatal variables, local-global temporal encoders, and patient-level statistical summaries through reliability-guided gating mechanisms, while jointly optimizing mortality prediction and an auxiliary length-of-stay objective. NeoTriFuse achieves competitive performance, with an F1 score of 0.6736 +/- 0.0216 and an AUROC of 0.9454 +/- 0.0056. Ablation studies indicate that the local-global temporal architecture and patient-level summary branch contribute most substantially to predictive performance, while reliability-aware gating provides additional improvements on threshold-dependent metrics under heterogeneous observation completeness. Sensitivity analyses further suggest stable performance across nearby hyperparameter settings. Overall, the findings support reliability-aware multimodal fusion as a practical approach for neonatal mortality prediction under realistic clinical missingness conditions.
Machine Learning (cs.LG)
arXiv:2608.26436 [cs.LG]
(or arXiv:2608.26436v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2608.26436 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 266 cuvinte
Subiecte conexe
A Confidence-Aware Multimodal Fusion Framework for Industrial Human-Robot Collaboration
arXiv · 10 sept. 2026
Context-aware Modality-Topology Co-Alignment for Multimodal Attributed Graphs
arXiv · 15 iun. 2026
RHEA: Reliability-Harmonized Reconstruction and Assignment for Robust Multimodal-Attributed Graph Clustering
arXiv · 4 aug. 2026
Source-Learned Reliance for Selective Test-Time Adaptation of Multimodal Time Series
arXiv · 7 oct. 2026