This paper presents a federated learning system for diabetic retinopathy detection. Mobile retinal scan data is processed locally across institutions, enabling collaborative model training while preserving patient data privacy and improving the robustness of clinical diagnoses in decentralized healthcare settings.

Charlotte Emmeline Foster, Hugo Sebastian Wainwright, Alice Rosamund Pike, Leonard Tobias Freeman, Amelia Georgia Newcombe

This study proposes an edge-deployed deep learning system for real-time monitoring of Parkinson’s disease progression. Using gait recognition from wearable sensors, the model supports continuous evaluation, aiding physicians in remote diagnosis and treatment optimization with reduced latency and bandwidth requirements.

Dominic Walter Holloway, Isla Francesca Kendall, Marcus Elliott Trafford, Fiona Isobel Prescott, Edward Nathaniel Byrne

This paper explores cross-modality image translation using GANs to fuse MRI and CT scans. The technique synthesizes complementary data representations, enhancing contrast and anatomical visibility, enabling improved segmentation, detection, and clinical interpretations across diverse imaging datasets in multimodal diagnostic systems.

Julian Albert Kingswood, Imogen Eloise Wallace, Nathaniel Charles Westridge, Clara Florence Seymour, Tobias Rupert Everhart

This study proposes a time-series-based predictive model for chronic kidney disease progression. It utilizes longitudinal EHR data and advanced machine learning algorithms to forecast stage transitions, enabling timely intervention and treatment planning tailored to individual patient trajectories.

Beatrice Helena Westmore, Oscar Patrick Brinsley, Matilda Grace Chandler, George Tristan Hollowell, Florence Isabel Ashcroft

This research presents a multi-agent reinforcement learning system designed to personalize healthcare pathways in smart hospitals. The model optimizes decisions on treatment sequences and resources, enhancing clinical efficiency and patient satisfaction through adaptive, AI-driven coordination among digital hospital subsystems.

Henry Maxwell Fairchild, Olivia Faith Ravenscroft, Benjamin Alistair Fairthorne, Charlotte Isabelle Fenwick, Rupert Elliott Mansfield

This study introduces an AI-driven virtual assistant designed for elderly post-operative patients. Utilizing speech-based natural language processing, it provides medication reminders, mobility suggestions, and symptom tracking, enhancing at-home recovery and reducing hospital readmissions in geriatric surgical populations.

Frederick Jameson Blythe, Georgia Madeleine Thornfield, Samuel Oliver Winslow, Lydia Abigail Penrose, Charles Xavier Redmond

This paper presents a hybrid framework for secure genomic data analysis. Blockchain ensures access control while homomorphic encryption preserves privacy in computations across distributed cloud environments, supporting secure, collaborative research on sensitive genomic datasets without compromising confidentiality or compliance.

Isabelle Margaret Winscott, Rupert Nathaniel Greystone, Eleanor Beatrix Lovelace, Gregory Thomas Lanchester, Victoria Grace Ainsworth