Home
Menu

Research

Developing and evaluating systems that connect medical images with language, diagnostic reasoning and structured clinical information.

Quantitative Clinical Imaging

Cephalometric landmark heatmap over a CBCT-derived radiograph.

Oral and Dental Image Analysis (ODIN 2026) · 2026

Automatic Cephalometric Landmark Localization on CBCT-Derived Digitally Reconstructed Radiographs for Skeletal Malocclusion Classification

Automated cephalometric landmark localisation on CBCT-derived radiographs and downstream skeletal malocclusion classification.

Explore project
Manual and automated ascites segmentations on axial CT images.

Radiology: Artificial Intelligence · 2024

Deep Learning Segmentation of Ascites on Abdominal CT Scans for Automatic Volume Quantification

Automated ascites segmentation and volume quantification across cirrhosis and ovarian-cancer CT cohorts.

Explore project
Annotated coronal CT body-composition segmentation.

International journal of computer assisted radiology and surgery · 2024

Enhanced muscle and fat segmentation for CT-based body composition analysis: a comparative study

Comparative evaluation of muscle and fat segmentation for CT-based body-composition analysis.

Explore project
Predicted ascites segmentation on a coronal CT image.

MedNeurIPS Workshop · 2022

Segmentation of Ascites on Abdominal CT Scans for the Assessment of Ovarian Cancer

Deep-learning segmentation of ascites on abdominal CT for quantitative assessment in ovarian cancer.

Explore project

Multimodal Clinical AI

Heatmap comparing multimodal model accuracy across radiology subspecialties.

Radiology · 2025

One year on: assessing progress of multimodal large language model performance on RSNA 2024 case of the day questions

A year-on-year evaluation of multimodal model performance on image-based RSNA diagnostic cases.

Explore project
Complete frontal chest radiograph from the RATCHET report example.

MICCAI (7) · 2021

RATCHET: Medical Transformer for Chest X-ray Diagnosis and Reporting

An end-to-end medical transformer for chest-radiograph diagnosis and report generation.

Explore project

Generative Medical Imaging

Synthesized retinal fundus images across diabetic-retinopathy grades.

Biomed. Opt. Express · 2023

High-fidelity diabetic retina fundus image synthesis from freestyle lesion maps

Controllable synthesis of retinal fundus images from generated or freehand diabetic-retinopathy lesion maps.

Explore project
Ground-truth and generated fetal-brain MRI slices at multiple positions.

MLMIR@MICCAI · 2019

Flexible Conditional Image Generation of Missing Data with Learned Mental Maps

Conditional generation of volumetric anatomy from sparse tomographic observations.

Explore project

Spatial Reasoning and Reconstruction

CNN pose-estimation architecture with a Riemannian loss.

MICCAI (1) · 2018

Computing CNN Loss and Gradients for Pose Estimation with Riemannian Geometry

Pose-estimation losses and gradients formulated directly on the rigid-transformation manifold.

Explore project
Fetal-brain reconstructions aligned to canonical atlas space.

IEEE Trans. Medical Imaging · 2018

3-D Reconstruction in Canonical Co-Ordinate Space From Arbitrarily Oriented 2-D Images

Learning canonical spatial coordinates for reconstructing volumes from arbitrarily oriented 2D images.

Explore project
Axial fetal-brain reconstruction produced by SVRNet.

MICCAI (2) · 2017

Predicting Slice-to-Volume Transformation in Presence of Arbitrary Subject Motion

Direct prediction of slice-to-volume transformations under substantial subject motion.

Explore project