Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder

Adam Marcus; Paul Bentley; Daniel Rueckert

doi:10.1007/978-3-031-44858-4_15

Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder

Adam Marcus
, Paul Bentley
, Daniel Rueckert

Informatics 31 - Chair of Artificial Intelligence in Healthcare and Medicine

Imperial College London

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Stroke is a major cause of death and disability worldwide. Accurate outcome and evolution prediction has the potential to revolutionize stroke care by individualizing clinical decision-making leading to better outcomes. However, despite a plethora of attempts and the rich data provided by neuroimaging, modelling the ultimate fate of brain tissue remains a challenging task. In this work, we apply recent ideas in the field of diffusion probabilistic models to generate a self-supervised semantically meaningful stroke representation from Computed Tomography (CT) images. We then improve this representation by extending the method to accommodate longitudinal images and the time from stroke onset. The effectiveness of our approach is evaluated on a dataset consisting of 5,824 CT images from 3,573 patients across two medical centers with minimal labels. Comparative experiments show that our method achieves the best performance for predicting next-day severity and functional outcome at discharge.

Original language	English
Title of host publication	Machine Learning in Clinical Neuroimaging - 6th International Workshop, MLCN 2023, Held in Conjunction with MICCAI 2023, Proceedings
Editors	Ahmed Abdulkadir, Deepti R. Bathula, Nicha C. Dvornek, Sindhuja T. Govindarajan, Mohamad Habes, Vinod Kumar, Esten Leonardsen, Thomas Wolfers, Yiming Xiao
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	162-153
Number of pages	10
ISBN (Print)	9783031448577
DOIs	https://doi.org/10.1007/978-3-031-44858-4_15
State	Published - 2023
Event	6th International Workshop on Machine Learning in Clinical Neuroimaging, MLCN 2023 - Vancouver, Canada Duration: 8 Oct 2023 → 12 Oct 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14312 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	6th International Workshop on Machine Learning in Clinical Neuroimaging, MLCN 2023
Country/Territory	Canada
City	Vancouver
Period	8/10/23 → 12/10/23

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Computed Tomography
Diffusion model
Outcome prediction
Stroke

Access to Document

10.1007/978-3-031-44858-4_15

Cite this

Marcus, A., Bentley, P., & Rueckert, D. (2023). Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder. In A. Abdulkadir, D. R. Bathula, N. C. Dvornek, S. T. Govindarajan, M. Habes, V. Kumar, E. Leonardsen, T. Wolfers, & Y. Xiao (Eds.), Machine Learning in Clinical Neuroimaging - 6th International Workshop, MLCN 2023, Held in Conjunction with MICCAI 2023, Proceedings (pp. 162-153). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14312 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-44858-4_15

Marcus, Adam ; Bentley, Paul ; Rueckert, Daniel. / Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder. Machine Learning in Clinical Neuroimaging - 6th International Workshop, MLCN 2023, Held in Conjunction with MICCAI 2023, Proceedings. editor / Ahmed Abdulkadir ; Deepti R. Bathula ; Nicha C. Dvornek ; Sindhuja T. Govindarajan ; Mohamad Habes ; Vinod Kumar ; Esten Leonardsen ; Thomas Wolfers ; Yiming Xiao. Springer Science and Business Media Deutschland GmbH, 2023. pp. 162-153 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0d9f6fbea27e4aecaf08c277ec4fcdea,

title = "Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder",

abstract = "Stroke is a major cause of death and disability worldwide. Accurate outcome and evolution prediction has the potential to revolutionize stroke care by individualizing clinical decision-making leading to better outcomes. However, despite a plethora of attempts and the rich data provided by neuroimaging, modelling the ultimate fate of brain tissue remains a challenging task. In this work, we apply recent ideas in the field of diffusion probabilistic models to generate a self-supervised semantically meaningful stroke representation from Computed Tomography (CT) images. We then improve this representation by extending the method to accommodate longitudinal images and the time from stroke onset. The effectiveness of our approach is evaluated on a dataset consisting of 5,824 CT images from 3,573 patients across two medical centers with minimal labels. Comparative experiments show that our method achieves the best performance for predicting next-day severity and functional outcome at discharge.",

keywords = "Computed Tomography, Diffusion model, Outcome prediction, Stroke",

author = "Adam Marcus and Paul Bentley and Daniel Rueckert",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.; 6th International Workshop on Machine Learning in Clinical Neuroimaging, MLCN 2023 ; Conference date: 08-10-2023 Through 12-10-2023",

year = "2023",

doi = "10.1007/978-3-031-44858-4\_15",

language = "English",

isbn = "9783031448577",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "162--153",

editor = "Ahmed Abdulkadir and Bathula, \{Deepti R.\} and Dvornek, \{Nicha C.\} and Govindarajan, \{Sindhuja T.\} and Mohamad Habes and Vinod Kumar and Esten Leonardsen and Thomas Wolfers and Yiming Xiao",

booktitle = "Machine Learning in Clinical Neuroimaging - 6th International Workshop, MLCN 2023, Held in Conjunction with MICCAI 2023, Proceedings",

}

Marcus, A, Bentley, P & Rueckert, D 2023, Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder. in A Abdulkadir, DR Bathula, NC Dvornek, ST Govindarajan, M Habes, V Kumar, E Leonardsen, T Wolfers & Y Xiao (eds), Machine Learning in Clinical Neuroimaging - 6th International Workshop, MLCN 2023, Held in Conjunction with MICCAI 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14312 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 162-153, 6th International Workshop on Machine Learning in Clinical Neuroimaging, MLCN 2023, Vancouver, Canada, 8/10/23. https://doi.org/10.1007/978-3-031-44858-4_15

Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder. / Marcus, Adam; Bentley, Paul; Rueckert, Daniel.
Machine Learning in Clinical Neuroimaging - 6th International Workshop, MLCN 2023, Held in Conjunction with MICCAI 2023, Proceedings. ed. / Ahmed Abdulkadir; Deepti R. Bathula; Nicha C. Dvornek; Sindhuja T. Govindarajan; Mohamad Habes; Vinod Kumar; Esten Leonardsen; Thomas Wolfers; Yiming Xiao. Springer Science and Business Media Deutschland GmbH, 2023. p. 162-153 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14312 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder

AU - Marcus, Adam

AU - Bentley, Paul

AU - Rueckert, Daniel

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

PY - 2023

Y1 - 2023

N2 - Stroke is a major cause of death and disability worldwide. Accurate outcome and evolution prediction has the potential to revolutionize stroke care by individualizing clinical decision-making leading to better outcomes. However, despite a plethora of attempts and the rich data provided by neuroimaging, modelling the ultimate fate of brain tissue remains a challenging task. In this work, we apply recent ideas in the field of diffusion probabilistic models to generate a self-supervised semantically meaningful stroke representation from Computed Tomography (CT) images. We then improve this representation by extending the method to accommodate longitudinal images and the time from stroke onset. The effectiveness of our approach is evaluated on a dataset consisting of 5,824 CT images from 3,573 patients across two medical centers with minimal labels. Comparative experiments show that our method achieves the best performance for predicting next-day severity and functional outcome at discharge.

AB - Stroke is a major cause of death and disability worldwide. Accurate outcome and evolution prediction has the potential to revolutionize stroke care by individualizing clinical decision-making leading to better outcomes. However, despite a plethora of attempts and the rich data provided by neuroimaging, modelling the ultimate fate of brain tissue remains a challenging task. In this work, we apply recent ideas in the field of diffusion probabilistic models to generate a self-supervised semantically meaningful stroke representation from Computed Tomography (CT) images. We then improve this representation by extending the method to accommodate longitudinal images and the time from stroke onset. The effectiveness of our approach is evaluated on a dataset consisting of 5,824 CT images from 3,573 patients across two medical centers with minimal labels. Comparative experiments show that our method achieves the best performance for predicting next-day severity and functional outcome at discharge.

KW - Computed Tomography

KW - Diffusion model

KW - Outcome prediction

KW - Stroke

UR - https://www.scopus.com/pages/publications/85191347543

U2 - 10.1007/978-3-031-44858-4_15

DO - 10.1007/978-3-031-44858-4_15

M3 - Conference contribution

AN - SCOPUS:85191347543

SN - 9783031448577

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 162

EP - 153

BT - Machine Learning in Clinical Neuroimaging - 6th International Workshop, MLCN 2023, Held in Conjunction with MICCAI 2023, Proceedings

A2 - Abdulkadir, Ahmed

A2 - Bathula, Deepti R.

A2 - Dvornek, Nicha C.

A2 - Govindarajan, Sindhuja T.

A2 - Habes, Mohamad

A2 - Kumar, Vinod

A2 - Leonardsen, Esten

A2 - Wolfers, Thomas

A2 - Xiao, Yiming

PB - Springer Science and Business Media Deutschland GmbH

T2 - 6th International Workshop on Machine Learning in Clinical Neuroimaging, MLCN 2023

Y2 - 8 October 2023 through 12 October 2023

ER -

Marcus A, Bentley P, Rueckert D. Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder. In Abdulkadir A, Bathula DR, Dvornek NC, Govindarajan ST, Habes M, Kumar V, Leonardsen E, Wolfers T, Xiao Y, editors, Machine Learning in Clinical Neuroimaging - 6th International Workshop, MLCN 2023, Held in Conjunction with MICCAI 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2023. p. 162-153. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-44858-4_15

Stroke Outcome and Evolution Prediction from CT Brain Using a Spatiotemporal Diffusion Autoencoder

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this