Project Contacts

For further information about this project, please contact:

• Dr Jie Ma (jie.ma2@newcastle.ac.uk)
• Dr Wanqing Zhao (wanqing.zhao@newcastle.ac.uk)

Project Description

This project develops and applies generative AI methods to longitudinal multimodal data obtained through microsampling blood data. The goal is to support early disease detection and explainable results by analysing data.

Software or Compute Methods

The work uses deep learning and generative modelling frameworks, primarily PyTorch with CUDA acceleration, alongside the standard Python scientific computing stack (NumPy, pandas, scikit-learn). Model architectures include variational autoencoders (VAEs) and diffusion-based generative models for learning latent representations of high-dimensional biomarker data.