Malignant Mesothelioma subtyping of tissue images via sampling driven multiple instance prediction
Conference paper
Eastwood, M., Marc, S., Gao, X., Sailem, H., Offman, J., Karteris, E., Montero Fernandez, A., Jonigk, D., Cookson, W., Moffatt, M., Popat, S., Minhas, F. and Robertus, J. 2022. Malignant Mesothelioma subtyping of tissue images via sampling driven multiple instance prediction. Michalowski, M., Abidi, S. and Abidi, S. (ed.) 20th International Conference on Artificial Intelligence in Medicine. Halifax, Canada 14 - 17 Jun 2022 Springer. https://doi.org/10.1007/978-3-031-09342-5_25
Type | Conference paper |
---|---|
Title | Malignant Mesothelioma subtyping of tissue images via sampling driven multiple instance prediction |
Authors | Eastwood, M., Marc, S., Gao, X., Sailem, H., Offman, J., Karteris, E., Montero Fernandez, A., Jonigk, D., Cookson, W., Moffatt, M., Popat, S., Minhas, F. and Robertus, J. |
Abstract | Malignant Mesothelioma is a difficult to diagnose and highly lethal cancer usually associated with asbestos exposure. It can be broadly classified into three subtypes: Epitheliod, Sarcomatoid, and Biphasic. Early diagnosis and identification of the subtype informs treatment and can help improve patient outcome. However, the subtyping of malignant mesothelioma, and specifically the recognition of transitional features from routine histology slides has a high level of inter-observer variability. In this work, we propose the first end-to-end multiple instance learning (MIL) approach for malignant mesothelioma subtyping. This uses an instance-based sampling scheme for training deep convolutional neural networks on this task that allows learning on a wider range of relevant instances compared to max or top-N based MIL approaches. The proposed MIL approach enables identification of malignant mesothelial sub-types of specific tissue regions. From this a continuous characterization of a sample according to predominance of sarcomatoid vs epithelioid regions is possible, thus avoiding the arbitrary and highly subjective categorisation by currently used subtypes. Instance scoring also enables studying tumor heterogeneity and identifying patterns associated with different subtypes. We have evaluated the proposed method on a dataset of 243 tissue micro-array cores with an AUROC of 0.87 ± 0.04 for this task. The dataset and developed methodology is available for the community at: https://github.com/measty/PINS. |
Keywords | Malignant mesothelioma; Multiple instance learning; Computational pathology ; Deep learning |
Sustainable Development Goals | 3 Good health and well-being |
Middlesex University Theme | Health & Wellbeing |
Research Group | Artificial Intelligence group |
Conference | 20th International Conference on Artificial Intelligence in Medicine |
Proceedings Title | Artificial Intelligence in Medicine: 20th International Conference on Artificial Intelligence in Medicine, AIME 2022, Halifax, NS, Canada, June 14–17, 2022, Proceedings |
Series | Lecture Notes in Computer Science |
Editors | Michalowski, M., Abidi, S. and Abidi, S. |
ISSN | 0302-9743 |
Electronic | 1611-3349 |
ISBN | |
Paperback | 9783031093418 |
Electronic | 9783031093425 |
Publisher | Springer |
Publication dates | |
Online | 09 Jul 2022 |
17 Jun 2022 | |
Publication process dates | |
Deposited | 30 Aug 2022 |
Accepted | 01 May 2022 |
Output status | Published |
Digital Object Identifier (DOI) | https://doi.org/10.1007/978-3-031-09342-5_25 |
Scopus EID | 2-s2.0-85135087810 |
Web of Science identifier | WOS:000878470100025 |
Language | English |
https://repository.mdx.ac.uk/item/89y4q
65
total views0
total downloads0
views this month0
downloads this month