This study proposes a hybrid deep learning system integrating dermoscopy images and clinical records to detect melanoma. The model enhances diagnostic accuracy through multi-source fusion, enabling dermatologists to identify malignant lesions early with greater confidence and reduced dependence on invasive procedures.

Jonathan Miles Whitaker, Hannah Elise Worthington, Felix Gregory Blake, Molly Ruthanne Carter, Sebastian Jude Thornton

This research presents an AI-powered predictive engine designed for real-time blood glucose monitoring. It leverages temporal patterns and physiological signals to forecast glycemic trends, enhancing patient management and reducing complications related to insulin dosing and lifestyle fluctuations in diabetes care.

Charlotte Ivy Manning, Edward Isaac Franklin, Olivia Mae Davidson, Julian Thomas Fletcher, Clara Abigail Newton

This paper introduces a lightweight ensemble classifier optimized to detect rare genetic conditions from neonatal health records. It utilizes stratified feature selection and hybrid decision pathways to improve early detection accuracy, ensuring better patient outcomes and timely specialized interventions in pediatrics.

Isaac Benjamin Ford, Emily Grace Whitmore, Adam Lawrence Cummings, Rebecca Louise Sinclair, Nathan Joel Hamilton

This study introduces a personalized reinforcement learning model for chemotherapy dosage optimization. The system adapts to patient-specific responses and biomarker feedback, enhancing therapeutic outcomes while minimizing adverse effects, thus aiding oncologists in designing safer, more efficient, and patient-centric cancer care protocols.

William Peter Sanderson, Lucy Harper Benson, Henry Maxwell Gibson, Grace Isabelle Carr, Thomas Finley Moore

This paper presents a smart AI pipeline deployed on edge devices to identify retinal pathologies in real-time. The system performs localized processing of fundus images, enabling affordable and accessible ocular diagnostics in remote or under-resourced environments with minimal computational overhead.

George Adrian Wallace, Amelia Frances Blake, Lucas Raymond Stevenson, Eleanor Grace Redford, Noah Douglas Whitman

This research proposes an AI-powered cardiac monitoring framework integrating multi-sensor wearables and time-series modeling to predict arrhythmias. The system enables continuous monitoring and early intervention strategies, significantly enhancing preventive healthcare for individuals at risk of cardiovascular complications.

Andrew Timothy Reynolds, Madeline Eve Collins, Oliver James Westbrook, Sophie Anne Hamilton, Joshua Elijah Porter

This paper introduces a machine learning framework designed to classify lung nodules from low-dose CT images. It improves early-stage lung cancer detection accuracy, offering radiologists a reliable decision-support tool that reduces interpretation time and enhances diagnostic confidence in oncology imaging workflows.

Benjamin Lucas Harper, Natalie Grace Ellis, Samuel David Fletcher, Zoe Katherine Harding, Ethan Daniel Montgomery