Data & Evidence Analysis
Reproducible R pipeline for Post-Market Clinical Follow-up (PMCF) studies under EU MDR. Spans the full chain from EDC-system data exports — cleaning, CDISC SDTM/ADaM mapping, and statistical modeling through to automated Tables, Figures and Listings (TFLs). Modeling covers mixed-effects/hierarchical, longitudinal (MMRM), survival, and Bayesian hierarchical approaches, with estimand-aligned imputation and sensitivity analyses. Reporting is orchestrated in Quarto for fully reproducible, audit-ready outputs that regenerate end-to-end as clinical data evolves.
Mathematical Modeling
Computational modeling of action potential propagation, reentry mechanisms, and arrhythmia substrates in mouse cardiac tissue. Implemented using openCARP with Hodgkin–Huxley and Markov-state ion channel formalisms to simulate electrophysiological behaviour and predict arrhythmia vulnerability.
Spatial Transcriptomics
Spatial transcriptomics analysis of human cardiac tissue on the 10x Visium platform. Processed Visium data through Space Ranger, performed QC and normalization in Seurat, and deconvolved spatial spots against a single-nucleus RNA-seq reference to estimate per-spot cell-type composition. Mapped spatially-resolved molecular niches across the myocardium, localizing distinct cell states and transcriptional programs within intact tissue architecture.
Experimental & Analysis
End-to-end experimental and analysis pipeline for optical mapping of cardiac electrophysiology on Langendorff-perfused mouse hearts. Covers voltage- and calcium-sensitive dye imaging, signal acquisition, AP/CaT extraction, conduction velocity mapping, and spatiotemporal visualization.