This study leverages genomic sequencing with deep neural networks to identify markers for metabolic disorders. Results show improved prediction accuracy and reliability in clinical scenarios, particularly during neonatal screenings, offering a non-invasive alternative to traditional metabolic profiling techniques.

David Raymond Foster, Li Wei Zhang, Ethan Michael Reynolds, Sarah Louise Peterson, Angela Renee Holloway

We propose a stratification model trained on electronic records to predict cardiac events. The approach improves patient outcomes and triage effectiveness by learning long-term dependencies in structured and unstructured medical histories from diverse populations across healthcare systems.

George Allan Winters, Maria Emilia Delgado, Kevin Zhou, Joshua David Clancy, Thomas Edward Walsh

This study introduces a machine learning model that integrates imaging and patient history to track kidney disease progression. The hybrid framework offers improved clinical decision support by merging renal ultrasound patterns with lab metrics to predict future deterioration risk.

Rachel Margaret Connors, Li Wei Huang, Frederick James O’Connor, Olivia Hannah Wells, Jean Baptiste Fournier

We introduce a federated deep learning framework that maintains data privacy while training on diagnostic images across hospitals. Our architecture ensures collaborative model accuracy without sharing raw patient data, adhering to strict clinical data protection regulations globally.

Leonard Charles Milton, Song Tao Chen, Hannah Victoria Briggs, Robert Jacob Dunn, Elias Dominic Fisher

This paper presents an explainable neural architecture for detecting and interpreting anomalies in pediatric brain MRIs. The model integrates saliency maps and attention layers to offer clinicians insight into decision pathways during neurodevelopmental disorder diagnosis.

Catherine Louise Brooks, Meilin Zhu, Marcus Phillip Anderson, Fiona Anne Davidson, Julian Lee Carter

This work explores the use of RNNs on ICU telemetry data for early sepsis warning. Our sequence-based classifier processes minute-level vitals, capturing onset trends that precede clinical symptoms and enabling earlier intervention protocols in critical care units.

Benjamin Harold Tucker, Alina Sophie Weber, Ricardo Manuel Torres, Emily Lauren Sharp, Samuel Vincent Doyle