Research
My PhD forecast children's vital signs in intensive care from bedside-monitor data. Before that, I helped build a model that predicts which human genes can be targeted by drugs.
PhD Thesis at UCL & Great Ormond Street Hospital
Multivariate ICU Forecasting with StemGNN
~12.5MObservationssliding-window segmented
4,120Raw admissionsbefore filtering
364Patients
5 sMonitor resolutionpaediatric ICU
I built the full clinical time-series pipeline: filtering from 4,120 raw admissions to 364 high-quality patients, age-aware normalisation, missing-data handling, outlier detection, and sliding-window segmentation across ~12.5M observations. A PyTorch StemGNN model then jointly forecasts heart rate and blood pressure from the 5-second bedside-monitor data, modelling the dependencies between them as a learned graph.
Journal Articles
Conference Presentations
Poster presentation at the IEEE Engineering in Medicine and Biology Society (EMBS) Conference2024
Poster presentation at the IEEE International Conference on Biomedical and Health Informatics (BHI)2024
Dissertations
Unsupervised ML on time-series data for clustering patient subtypes and improving decision supportMRes in AI-enabled Healthcare, University College London, 2020 – 2021
DrugnomeAI – an ensemble framework for predicting druggability of human genes in collaboration with AstraZenecaMSc in Data Science, Lancaster University, 2019 – 2020