Biography
Dr Abhirup Banerjee is a Royal Society University Research Fellow and Principal Investigator at the Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford. His research lies at the intersection of cardiovascular science, artificial intelligence, and computational modelling, with a focus on digital twins, geometric machine learning, and multimodal data integration for cardiac diagnostics and interventions.
He leads the Multimodal Medical Data Integration & Analysis (MultiMeDIA) Lab, where his team develops personalised, predictive models of cardiac anatomy and function using large-scale imaging and physiological datasets. A key aspect of his work involves reconstructing patient-specific 3D/4D cardiac structures from coronary angiography, cardiac MRI, and ECG. His AI-driven pipelines for coronary reconstruction, infarction modelling, and atrial fibrillation mapping are designed for real-time clinical use and have been patented in collaboration with Oxford University Innovation.
Dr Banerjee’s approach to cardiovascular science is rooted in the application of advanced computational methodologies, including variational autoencoders, point cloud networks, graph-based attention models, and statistical shape analysis. His work exemplifies a commitment to interdisciplinary innovation, translating cutting-edge algorithms into clinically meaningful tools.
He has authored over 80 peer-reviewed publications and serves on the Editorial boards of several international journals. His research has been widely presented at leading scientific meetings, contributing to the advancement of data-driven approaches in cardiovascular medicine. He is also actively engaged in public outreach, regularly participating in science exhibitions, open days, and community engagement events to promote awareness of biomedical engineering and digital health.
Awards and Honours
The Young Scientist Award from the Indian Science Congress Association in 2017 in the Section of Information and Communication Science & Technology (including Computer Sciences).
Most Recent Publications
Contrastive Machine Learning to Quantify Hypertensive Multiorgan Damage and Identify New Disease Phenotypes: A Multinational Multimodal Study.
Contrastive Machine Learning to Quantify Hypertensive Multiorgan Damage and Identify New Disease Phenotypes: A Multinational Multimodal Study.
Coronary Arteries Segmentation in Invasive X-ray Angiography: A Comprehensive Review and Benchmarking
Coronary Arteries Segmentation in Invasive X-ray Angiography: A Comprehensive Review and Benchmarking
Histo-Unet: Histopathology Image Segmentation Using Topology-Aware Unet with Dual Uncertainty Quantification
Histo-Unet: Histopathology Image Segmentation Using Topology-Aware Unet with Dual Uncertainty Quantification
Multi-Class Transcompletion: Geometric Deep Learning for Four-Chamber Cardiac Reconstruction From Cine MRI
Multi-Class Transcompletion: Geometric Deep Learning for Four-Chamber Cardiac Reconstruction From Cine MRI
Feature Invariance Via Interpretable Ablation for Single-Source Domain Generalization in X-Ray Angiography Segmentation
Feature Invariance Via Interpretable Ablation for Single-Source Domain Generalization in X-Ray Angiography Segmentation
Research Interests
Dr Banerjee’s research interests are primarily focused on (but not limited to) Multimodal Reconstruction and Analysis of Digital Anatomy for Real-Time Clinical Interventions; in particular Segmentation, Landmark Detection, Tracking, Motion Modelling, Reconstruction, Registration, and Fusion of 2D/3D/3D+t Human Anatomical Structures (Cardiovascular, Cerebral, Biliary System, etc.) and Physiological information from Multi-Modality including X-ray, Angiography, MRI, CT, US, etc.
The video below showcases some initial research imapcts of Dr Banerjee’s work for patients and the NHS.
Current Projects
- Multimodal Reconstruction of Digital Heart for Cardiac Interventions in Real-Time.
- Building Personalised Heart and Torso Models from Clinical MRI Scans for Simulation of Cardiac Electromechanics.
- Rough-Probabilistic Modelling for Brain MRI Analysis, etc.
Research Groups
Most Recent Publications
Contrastive Machine Learning to Quantify Hypertensive Multiorgan Damage and Identify New Disease Phenotypes: A Multinational Multimodal Study.
Contrastive Machine Learning to Quantify Hypertensive Multiorgan Damage and Identify New Disease Phenotypes: A Multinational Multimodal Study.
Coronary Arteries Segmentation in Invasive X-ray Angiography: A Comprehensive Review and Benchmarking
Coronary Arteries Segmentation in Invasive X-ray Angiography: A Comprehensive Review and Benchmarking
Histo-Unet: Histopathology Image Segmentation Using Topology-Aware Unet with Dual Uncertainty Quantification
Histo-Unet: Histopathology Image Segmentation Using Topology-Aware Unet with Dual Uncertainty Quantification
Multi-Class Transcompletion: Geometric Deep Learning for Four-Chamber Cardiac Reconstruction From Cine MRI
Multi-Class Transcompletion: Geometric Deep Learning for Four-Chamber Cardiac Reconstruction From Cine MRI
Feature Invariance Via Interpretable Ablation for Single-Source Domain Generalization in X-Ray Angiography Segmentation
Feature Invariance Via Interpretable Ablation for Single-Source Domain Generalization in X-Ray Angiography Segmentation
DPhil & Postdoc Opportunities
If you are passionate about embarking on a journey together on solving Real-life Biomedical problems with Clinical Impacts, please contact me via Email, LinkedIn, or Twitter.
At the moment, I have openings for both DPhil students and Postdoctoral Researchers in my Team.
Most Recent Publications
Contrastive Machine Learning to Quantify Hypertensive Multiorgan Damage and Identify New Disease Phenotypes: A Multinational Multimodal Study.
Contrastive Machine Learning to Quantify Hypertensive Multiorgan Damage and Identify New Disease Phenotypes: A Multinational Multimodal Study.
Coronary Arteries Segmentation in Invasive X-ray Angiography: A Comprehensive Review and Benchmarking
Coronary Arteries Segmentation in Invasive X-ray Angiography: A Comprehensive Review and Benchmarking
Histo-Unet: Histopathology Image Segmentation Using Topology-Aware Unet with Dual Uncertainty Quantification
Histo-Unet: Histopathology Image Segmentation Using Topology-Aware Unet with Dual Uncertainty Quantification
Multi-Class Transcompletion: Geometric Deep Learning for Four-Chamber Cardiac Reconstruction From Cine MRI
Multi-Class Transcompletion: Geometric Deep Learning for Four-Chamber Cardiac Reconstruction From Cine MRI
Feature Invariance Via Interpretable Ablation for Single-Source Domain Generalization in X-Ray Angiography Segmentation
Feature Invariance Via Interpretable Ablation for Single-Source Domain Generalization in X-Ray Angiography Segmentation
