s-SuStaIn: Scaling subtype and stage inference via simultaneous clustering of subjects and biomarkers

Raghav Tandon, James J Lah, Cassie S Mitchell
Proceedings of the fifth Conference on Health, Inference, and Learning, PMLR 248:461-476, 2024.

Abstract

Event-based models (EBM) provide an important platform for modeling disease progression. This work successfully extends previous EBM approaches to work with larger sets of biomarkers while simultaneously modeling heterogeneity in disease progression trajectories. We develop and validate the s-SuStain method for scalable event-based modeling of disease progression subtypes using large numbers of features. s-SuStaIn is typically an order of magnitude faster than its predecessor (SuStaIn). Moreover, we perform a case study with s-SuStaIn using open access cross-sectional Alzheimer’s Disease Neuroimaging (ADNI) data to stage AD patients into four subtypes based on dynamic disease progression. s-SuStaIn shows that the inferred subtypes and stages predict progression to AD among MCI subjects. The subtypes show difference in AD incidence-rates and reveal clinically meaningful progression trajectories when mapped to a brain atlas.

Cite this Paper

BibTeX
@InProceedings{pmlr-v248-tandon24a, title = {s-SuStaIn: Scaling subtype and stage inference via simultaneous clustering of subjects and biomarkers}, author = {Tandon, Raghav and Lah, James J and Mitchell, Cassie S}, booktitle = {Proceedings of the fifth Conference on Health, Inference, and Learning}, pages = {461--476}, year = {2024}, editor = {Pollard, Tom and Choi, Edward and Singhal, Pankhuri and Hughes, Michael and Sizikova, Elena and Mortazavi, Bobak and Chen, Irene and Wang, Fei and Sarker, Tasmie and McDermott, Matthew and Ghassemi, Marzyeh}, volume = {248}, series = {Proceedings of Machine Learning Research}, month = {27--28 Jun}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v248/main/assets/tandon24a/tandon24a.pdf}, url = {https://proceedings.mlr.press/v248/tandon24a.html}, abstract = {Event-based models (EBM) provide an important platform for modeling disease progression. This work successfully extends previous EBM approaches to work with larger sets of biomarkers while simultaneously modeling heterogeneity in disease progression trajectories. We develop and validate the s-SuStain method for scalable event-based modeling of disease progression subtypes using large numbers of features. s-SuStaIn is typically an order of magnitude faster than its predecessor (SuStaIn). Moreover, we perform a case study with s-SuStaIn using open access cross-sectional Alzheimer’s Disease Neuroimaging (ADNI) data to stage AD patients into four subtypes based on dynamic disease progression. s-SuStaIn shows that the inferred subtypes and stages predict progression to AD among MCI subjects. The subtypes show difference in AD incidence-rates and reveal clinically meaningful progression trajectories when mapped to a brain atlas.} }
Endnote
%0 Conference Paper %T s-SuStaIn: Scaling subtype and stage inference via simultaneous clustering of subjects and biomarkers %A Raghav Tandon %A James J Lah %A Cassie S Mitchell %B Proceedings of the fifth Conference on Health, Inference, and Learning %C Proceedings of Machine Learning Research %D 2024 %E Tom Pollard %E Edward Choi %E Pankhuri Singhal %E Michael Hughes %E Elena Sizikova %E Bobak Mortazavi %E Irene Chen %E Fei Wang %E Tasmie Sarker %E Matthew McDermott %E Marzyeh Ghassemi %F pmlr-v248-tandon24a %I PMLR %P 461--476 %U https://proceedings.mlr.press/v248/tandon24a.html %V 248 %X Event-based models (EBM) provide an important platform for modeling disease progression. This work successfully extends previous EBM approaches to work with larger sets of biomarkers while simultaneously modeling heterogeneity in disease progression trajectories. We develop and validate the s-SuStain method for scalable event-based modeling of disease progression subtypes using large numbers of features. s-SuStaIn is typically an order of magnitude faster than its predecessor (SuStaIn). Moreover, we perform a case study with s-SuStaIn using open access cross-sectional Alzheimer’s Disease Neuroimaging (ADNI) data to stage AD patients into four subtypes based on dynamic disease progression. s-SuStaIn shows that the inferred subtypes and stages predict progression to AD among MCI subjects. The subtypes show difference in AD incidence-rates and reveal clinically meaningful progression trajectories when mapped to a brain atlas.
APA
Tandon, R., Lah, J.J. & Mitchell, C.S.. (2024). s-SuStaIn: Scaling subtype and stage inference via simultaneous clustering of subjects and biomarkers. Proceedings of the fifth Conference on Health, Inference, and Learning, in Proceedings of Machine Learning Research 248:461-476 Available from https://proceedings.mlr.press/v248/tandon24a.html.

Related Material