Predicting post-operative right ventricular failure using video-based deep learning

Rohan Shad; Nicolas Quach; Robyn Fong; Patpilai Kasinpila; Cayley Bowles; Miguel Castro; Ashrith Guha; Erik E. Suarez; Stefan Jovinge; Sangjin Lee; Theodore Boeve; Myriam Amsallem; Xiu Tang; Francois Haddad; Yasuhiro Shudo; Y. Joseph Woo; Jeffrey Teuteberg; John P. Cunningham; Curtis P. Langlotz; William Hiesinger

doi:10.1038/s41467-021-25503-9

Predicting post-operative right ventricular failure using video-based deep learning

Rohan Shad, Nicolas Quach, Robyn Fong, Patpilai Kasinpila, Cayley Bowles, Miguel Castro, Ashrith Guha, Erik E. Suarez, Stefan Jovinge, Sangjin Lee, Theodore Boeve, Myriam Amsallem, Xiu Tang, Francois Haddad, Yasuhiro Shudo, Y. Joseph Woo, Jeffrey Teuteberg, John P. Cunningham, Curtis P. Langlotz, William Hiesinger

Research output: Contribution to journal › Article › peer-review

34 Scopus citations

Abstract

Despite progressive improvements over the decades, the rich temporally resolved data in an echocardiogram remain underutilized. Human assessments reduce the complex patterns of cardiac wall motion, to a small list of measurements of heart function. All modern echocardiography artificial intelligence (AI) systems are similarly limited by design – automating measurements of the same reductionist metrics rather than utilizing the embedded wealth of data. This underutilization is most evident where clinical decision making is guided by subjective assessments of disease acuity. Predicting the likelihood of developing post-operative right ventricular failure (RV failure) in the setting of mechanical circulatory support is one such example. Here we describe a video AI system trained to predict post-operative RV failure using the full spatiotemporal density of information in pre-operative echocardiography. We achieve an AUC of 0.729, and show that this ML system significantly outperforms a team of human experts at the same task on independent evaluation.

Original language	English (US)
Article number	5192
Journal	Nature Communications
Volume	12
Issue number	1
DOIs	https://doi.org/10.1038/s41467-021-25503-9
State	Published - Dec 1 2021

ASJC Scopus subject areas

Chemistry(all)
Biochemistry, Genetics and Molecular Biology(all)
General
Physics and Astronomy(all)

Access to Document

10.1038/s41467-021-25503-9

Cite this

Shad, R., Quach, N., Fong, R., Kasinpila, P., Bowles, C., Castro, M., Guha, A., Suarez, E. E., Jovinge, S., Lee, S., Boeve, T., Amsallem, M., Tang, X., Haddad, F., Shudo, Y., Woo, Y. J., Teuteberg, J., Cunningham, J. P., Langlotz, C. P., & Hiesinger, W. (2021). Predicting post-operative right ventricular failure using video-based deep learning. Nature Communications, 12(1), Article 5192. https://doi.org/10.1038/s41467-021-25503-9

Shad, R, Quach, N, Fong, R, Kasinpila, P, Bowles, C, Castro, M, Guha, A , Suarez, EE, Jovinge, S, Lee, S, Boeve, T, Amsallem, M, Tang, X, Haddad, F, Shudo, Y, Woo, YJ, Teuteberg, J, Cunningham, JP, Langlotz, CP & Hiesinger, W 2021, 'Predicting post-operative right ventricular failure using video-based deep learning', Nature Communications, vol. 12, no. 1, 5192. https://doi.org/10.1038/s41467-021-25503-9

@article{4874014f0a2f4a58923fdc50efcc589c,

title = "Predicting post-operative right ventricular failure using video-based deep learning",

abstract = "Despite progressive improvements over the decades, the rich temporally resolved data in an echocardiogram remain underutilized. Human assessments reduce the complex patterns of cardiac wall motion, to a small list of measurements of heart function. All modern echocardiography artificial intelligence (AI) systems are similarly limited by design – automating measurements of the same reductionist metrics rather than utilizing the embedded wealth of data. This underutilization is most evident where clinical decision making is guided by subjective assessments of disease acuity. Predicting the likelihood of developing post-operative right ventricular failure (RV failure) in the setting of mechanical circulatory support is one such example. Here we describe a video AI system trained to predict post-operative RV failure using the full spatiotemporal density of information in pre-operative echocardiography. We achieve an AUC of 0.729, and show that this ML system significantly outperforms a team of human experts at the same task on independent evaluation.",

author = "Rohan Shad and Nicolas Quach and Robyn Fong and Patpilai Kasinpila and Cayley Bowles and Miguel Castro and Ashrith Guha and Suarez, {Erik E.} and Stefan Jovinge and Sangjin Lee and Theodore Boeve and Myriam Amsallem and Xiu Tang and Francois Haddad and Yasuhiro Shudo and Woo, {Y. Joseph} and Jeffrey Teuteberg and Cunningham, {John P.} and Langlotz, {Curtis P.} and William Hiesinger",

note = "Funding Information: Some of the computing for this project was performed on the Stanford Sherlock and Nero clusters. We would like to thank Stanford University and the Stanford Research Computing Center for providing computational resources and support. This project was in part supported by a Stanford Artificial Intelligence in Medicine and Imaging (AIMI) Seed Grant, and Cloud Compute Credits from Google Cloud. R.S. was supported in part by the American Heart Association Postdoctoral Fellowship Award (Grant #834986). Publisher Copyright: {\textcopyright} 2021, The Author(s).",

year = "2021",

month = dec,

day = "1",

doi = "10.1038/s41467-021-25503-9",

language = "English (US)",

volume = "12",

journal = "Nature Communications",

issn = "2041-1723",

publisher = "Nature Publishing Group",

number = "1",

}

TY - JOUR

T1 - Predicting post-operative right ventricular failure using video-based deep learning

AU - Shad, Rohan

AU - Quach, Nicolas

AU - Fong, Robyn

AU - Kasinpila, Patpilai

AU - Bowles, Cayley

AU - Castro, Miguel

AU - Guha, Ashrith

AU - Suarez, Erik E.

AU - Jovinge, Stefan

AU - Lee, Sangjin

AU - Boeve, Theodore

AU - Amsallem, Myriam

AU - Tang, Xiu

AU - Haddad, Francois

AU - Shudo, Yasuhiro

AU - Woo, Y. Joseph

AU - Teuteberg, Jeffrey

AU - Cunningham, John P.

AU - Langlotz, Curtis P.

AU - Hiesinger, William

N1 - Funding Information: Some of the computing for this project was performed on the Stanford Sherlock and Nero clusters. We would like to thank Stanford University and the Stanford Research Computing Center for providing computational resources and support. This project was in part supported by a Stanford Artificial Intelligence in Medicine and Imaging (AIMI) Seed Grant, and Cloud Compute Credits from Google Cloud. R.S. was supported in part by the American Heart Association Postdoctoral Fellowship Award (Grant #834986). Publisher Copyright: © 2021, The Author(s).

PY - 2021/12/1

Y1 - 2021/12/1

N2 - Despite progressive improvements over the decades, the rich temporally resolved data in an echocardiogram remain underutilized. Human assessments reduce the complex patterns of cardiac wall motion, to a small list of measurements of heart function. All modern echocardiography artificial intelligence (AI) systems are similarly limited by design – automating measurements of the same reductionist metrics rather than utilizing the embedded wealth of data. This underutilization is most evident where clinical decision making is guided by subjective assessments of disease acuity. Predicting the likelihood of developing post-operative right ventricular failure (RV failure) in the setting of mechanical circulatory support is one such example. Here we describe a video AI system trained to predict post-operative RV failure using the full spatiotemporal density of information in pre-operative echocardiography. We achieve an AUC of 0.729, and show that this ML system significantly outperforms a team of human experts at the same task on independent evaluation.

AB - Despite progressive improvements over the decades, the rich temporally resolved data in an echocardiogram remain underutilized. Human assessments reduce the complex patterns of cardiac wall motion, to a small list of measurements of heart function. All modern echocardiography artificial intelligence (AI) systems are similarly limited by design – automating measurements of the same reductionist metrics rather than utilizing the embedded wealth of data. This underutilization is most evident where clinical decision making is guided by subjective assessments of disease acuity. Predicting the likelihood of developing post-operative right ventricular failure (RV failure) in the setting of mechanical circulatory support is one such example. Here we describe a video AI system trained to predict post-operative RV failure using the full spatiotemporal density of information in pre-operative echocardiography. We achieve an AUC of 0.729, and show that this ML system significantly outperforms a team of human experts at the same task on independent evaluation.

UR - http://www.scopus.com/inward/record.url?scp=85114413399&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85114413399&partnerID=8YFLogxK

U2 - 10.1038/s41467-021-25503-9

DO - 10.1038/s41467-021-25503-9

M3 - Article

C2 - 34465780

AN - SCOPUS:85114413399

SN - 2041-1723

VL - 12

JO - Nature Communications

JF - Nature Communications

IS - 1

M1 - 5192

ER -

Predicting post-operative right ventricular failure using video-based deep learning

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this