collaborators

5 papers

cs.CV2026

Tokenizer Generator Coupling in Medical Image Generation

Liam Chalcroft

Latent medical image generators usually treat the tokenizer as fixed preprocessing. We test whether this separation is valid in a controlled ChestMNIST study at 64x64, crossing dis…

cs.LG2026

GAZE: Grounded Agentic Zero-shot Evaluation with Viewer-Level Tools and Literature Retrieval on Rare Brain MRI

Duaa Alim, Mogtaba Alim, Liam Chalcroft

Vision-language models (VLMs) read an image and produce text in a single forward pass, whereas radiologists typically inspect an image several times and consult the literature befo…

eess.IV2025

Synthetic Data for Robust Stroke Segmentation

Liam Chalcroft, Ioannis Pappas, Cathy J. Price +1

Current deep learning-based approaches to lesion segmentation in neuroimaging often depend on high-resolution images and extensive annotated data, limiting clinical applicability.…

cs.CV2025

Unified 3D MRI Representations via Sequence-Invariant Contrastive Learning

Liam Chalcroft, Jenny Crinion, Cathy J. Price +1

Self-supervised deep learning has accelerated 2D natural image analysis but remains difficult to translate into 3D MRI, where data are scarce and pre-trained 2D backbones cannot ca…

eess.IV2025

Domain-Agnostic Stroke Lesion Segmentation Using Physics-Constrained Synthetic Data

Liam Chalcroft, Jenny Crinion, Cathy J. Price +1

Segmenting stroke lesions in MRI is challenging due to diverse acquisition protocols that limit model generalisability. In this work, we introduce two physics-constrained approache…