Amaya Gallagher Syed

Amaya Gallagher Syed

My work focuses on multi-omic data integration, histopathology imaging and deep learning models applied to autoimmune diseases.

Publications

Going Beyond H&E and Oncology: How Do Histopathology Foundation Models Perform for Multi-stain IHC and Immunology?

Published in Workshop on Advancements In Medical Foundation Models @ NeurIPS24, 2024

Do histopathology Foundation Models trained on H&E cancer datasets enhance performance for Immunohistochemistry autoimmune datasets? Do they produce more relevant Whole Slide Image heatmaps, aligned with autoimmune aetiology? We empirically examine these questions in this paper.

Recommended citation: Amaya Gallagher-Syed, Elena Pontarini, Myles J. Lewis, Michael R. Barnes and Gregory Slabaugh, Going Beyond H&E and Oncology: How Do Histopathology Foundation Models Perform for Multi-stain IHC and Immunology?. Workshop on Advancements In Medical Foundation Models, NeurIPS 2024, Vancouver, Canada https://arxiv.org/abs/2410.21560

MUSTANG: Multi-Stain Self-Attention Graph Multiple Instance Learning Pipeline for Histopathology Whole Slide Images

Published in British Machine Vision Conference, 2023

Here we propose an end-to-end multi-stain self-attention graph (MUSTANG) multiple instance learning pipeline, which is designed to solve a weakly-supervised gigapixel multi-image classification task, where the label is assigned at the patient-level, but no slide-level labels or region annotations are available. The pipeline uses a self-attention based approach by restricting the operations to a highly sparse k-Nearest Neighbour Graph of embedded WSI patches based on the Euclidean distance.

Recommended citation: Gallagher-Syed, A., Rossi, L., Rivellese, F., Pitzalis, C., Lewis, M., Barnes, M. and Slabaugh, G., 2023. Multi-Stain Self-Attention Graph Multiple Instance Learning Pipeline for Histopathology Whole Slide Images. arXiv preprint arXiv:2309.10650. https://arxiv.org/abs/2309.10650.pdf

Automated segmentation of rheumatoid arthritis immunohistochemistry stained synovial tissue

Published in 27th Conference on Medical Image Understanding and Analysis, 2023

We train a UNET on a hand-curated, heterogeneous real-world multi-centre clinical dataset R4RA, which contains multiple types of IHC staining. The model obtains a DICE score of 0.865 and successfully segments different types of IHC staining, as well as dealing with variance in colours, intensity and common WSIs artefacts from the different clinical centres.

Recommended citation: Gallagher-Syed, A., Khan, A., Rivellese, F., Pitzalis, C., Lewis, M.J., Slabaugh, G. and Barnes, M.R., 2023. Automated segmentation of rheumatoid arthritis immunohistochemistry stained synovial tissue. arXiv preprint arXiv:2309.07255. https://arxiv.org/abs/2309.07255.pdf