Recent publications in Frontiers journals cover innovative approaches in fetal monitoring, health indicator frameworks, Alzheimer’s prediction, and more. Here are key studies published around early January 2026.
Machine Learning Model for Fetal Health Assessment During Pregnancy
Researchers developed a multimodal vibroacoustic sensor array combined with machine learning to detect fetal movements objectively. Tested against ultrasound in 25 pregnant participants, an ensemble RUSBoost model achieved precision of 0.44 and recall of 0.61 using concatenated piezoelectric and acoustic data.
This low-cost, wearable approach could enable continuous monitoring in low- and middle-income countries, where stillbirth rates remain high due to limited access to ultrasound. It addresses subjectivity in maternal perception and resource constraints in clinical settings.
FAIR Foundations for NCD Indicator Vault in the EU
A feasibility study proposes the SOLICIT metadata model to contextualize indicators for non-communicable diseases across EU countries. Adapted from generic frameworks and extended to diabetes and other NCDs, it applies FAIR principles for better comparability and interoperability.
The work includes seven recommendations for EU implementation, such as piloting on federated networks and developing user-friendly portals. Enhanced metadata could improve transparency, reduce bias, and support evidence-based policy.
Deep Multimodal Learning Predicts Domain-Level Cognitive Decline in Alzheimer’s
Using ADNI data from 653 participants, researchers built models integrating MRI, PET scans, neuropsychological tests, and demographics. CNN- and GNN-based architectures with pre-training explained 29–36% of variance in decline rates and achieved AUC >0.83 for detecting deterioration in memory, language, and executive domains.
This advances beyond stage-transition predictions toward precise, domain-specific trajectories—crucial for personalized care and trial enrichment in Alzheimer’s heterogeneous progression.
These open-access studies demonstrate practical AI and data innovations addressing global health challenges, from maternal care to neurodegeneration.
Sources:
- Ashik et al. (2025). Frontiers in Bioengineering and Biotechnology. DOI: 10.3389/fbioe.2025.1691064
- Štotl et al. (2025). Frontiers in Digital Health. DOI: 10.3389/fdgth.2025.1685733
- García-Gutiérrez et al. (2025). Frontiers in Artificial Intelligence. DOI: 10.3389/frai.2025.1731062
