Medical device AI
Machine learning systems for regulated medical environments: cardiac imaging algorithms, intravascular tissue characterisation, and bronchoscopic decision support under FDA and EU Medical Device Regulation (CE MDR) pathways.
Clinical study programmes: SEPARATE (in vivo peripheral arterial clot trial, 17 patients), E-SEPARATE (ex vivo thrombus validation versus histology, 15 patients), and INSPECT (first in human lung tissue classification during bronchoscopic biopsy, 26 patients).
Cardiac ultrasound AI
As Head of Research, I lead AI/ML for cardiac ultrasound (classification, segmentation, detection) under FDA Predetermined Change Control Plan (PCCP) and CE MDR pathways. I steer research direction, build validation documentation as milestones are reached, and coordinate across AI, clinical, and regulatory functions.
Clot characterisation in peripheral arterial disease
Applied multimodal ML to peripheral arterial disease using the Clotild® Smart Guidewire System. First author on two clinical trial abstract programmes (SEPARATE and E-SEPARATE), each presented at Paris Vascular Insights 2024 and JET OPEN the world 2025.
- SEPARATE: 17 patients; 100% primary endpoint success; impedance based identification of clot rich in red blood cells
- E-SEPARATE: 15 patients; R²=0.79 vs histology gold standard
Clotild® holds FDA Breakthrough Device designation.
JET OPEN 2025 abstracts (author listing) · SEPARATE trial registry · Publications
Lung tissue classification during bronchoscopic biopsy
First in human study of in situ lung tissue characterisation during bronchoscopic biopsy. ML analysis lead; third author on American Thoracic Society 2026 clinical abstract.
- 26 patients across Australia and France
- 80.9% accuracy for healthy vs lesion tissue
- Learning curves projecting above 90% with additional data
ATS 2026 session listing · ATS 2026 results (BioSpace) · INSPECT trial registry · Biosignal methods · Publications