IDEAS home Printed from https://ideas.repec.org/a/gam/jforec/v6y2024i2p22-417d1402397.html
   My bibliography  Save this article

Deep Survival Models Can Improve Long-Term Mortality Risk Estimates from Chest Radiographs

Author

Listed:
  • Mingzhu Liu

    (Auton Laboratory, Carnegie Mellon University, Pittsburgh, PA 15213, USA)

  • Chirag Nagpal

    (Auton Laboratory, Carnegie Mellon University, Pittsburgh, PA 15213, USA)

  • Artur Dubrawski

    (Auton Laboratory, Carnegie Mellon University, Pittsburgh, PA 15213, USA)

Abstract

Deep learning has recently demonstrated the ability to predict long-term patient risk and its stratification when trained on imaging data such as chest radiographs. However, existing methods formulate estimating patient risk as a binary classification, typically ignoring or limiting the use of temporal information, and not accounting for the loss of patient follow-up, which reduces the fidelity of estimation and limits the prediction to a certain time horizon. In this paper, we demonstrate that deep survival and time-to-event prediction models can outperform binary classifiers at predicting mortality and risk of adverse health events. In our study, deep survival models were trained to predict risk scores from chest radiographs and patient demographic information in the Prostate, Lung, Colorectal, and Ovarian (PLCO) cancer screening trial (25,433 patient data points used in this paper) for 2-, 5-, and 10-year time horizons. Binary classification models that predict mortality at these time horizons were built as baselines. Compared to the considered alternative, deep survival models improve the Brier score (5-year: 0.0455 [95% CI, 0.0427–0.0482] vs. 0.0555 [95% CI, (0.0535–0.0575)], p < 0.05) and expected calibration error (ECE) (5-year: 0.0110 [95% CI, 0.0080–0.0141] vs. 0.0747 [95% CI, 0.0718–0.0776], p < 0.05) for those fixed time horizons and are able to generate predictions for any time horizon, without the need to retrain the models. Our study suggests that deep survival analysis tools can outperform binary classification in terms of both discriminative performance and calibration, offering a potentially plausible solution for forecasting risk in clinical practice.

Suggested Citation

  • Mingzhu Liu & Chirag Nagpal & Artur Dubrawski, 2024. "Deep Survival Models Can Improve Long-Term Mortality Risk Estimates from Chest Radiographs," Forecasting, MDPI, vol. 6(2), pages 1-14, May.
  • Handle: RePEc:gam:jforec:v:6:y:2024:i:2:p:22-417:d:1402397
    as

    Download full text from publisher

    File URL: https://www.mdpi.com/2571-9394/6/2/22/pdf
    Download Restriction: no

    File URL: https://www.mdpi.com/2571-9394/6/2/22/
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Uno, Hajime & Cai, Tianxi & Tian, Lu & Wei, L.J., 2007. "Evaluating Prediction Rules for t-Year Survivors With Censored Regression Models," Journal of the American Statistical Association, American Statistical Association, vol. 102, pages 527-537, June.
    2. Pranav Rajpurkar & Jeremy Irvin & Robyn L Ball & Kaylie Zhu & Brandon Yang & Hershel Mehta & Tony Duan & Daisy Ding & Aarti Bagul & Curtis P Langlotz & Bhavik N Patel & Kristen W Yeom & Katie Shpanska, 2018. "Deep learning for chest radiograph diagnosis: A retrospective comparison of the CheXNeXt algorithm to practicing radiologists," PLOS Medicine, Public Library of Science, vol. 15(11), pages 1-17, November.
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Yu Zheng & Tianxi Cai, 2017. "Augmented estimation for t‐year survival with censored regression models," Biometrics, The International Biometric Society, vol. 73(4), pages 1169-1178, December.
    2. Paul Frédéric Blanche & Anders Holt & Thomas Scheike, 2023. "On logistic regression with right censored data, with or without competing risks, and its use for estimating treatment effects," Lifetime Data Analysis: An International Journal Devoted to Statistical Methods and Applications for Time-to-Event Data, Springer, vol. 29(2), pages 441-482, April.
    3. Eric Engle & Andrei Gabrielian & Alyssa Long & Darrell E Hurt & Alex Rosenthal, 2020. "Performance of Qure.ai automatic classifiers against a large annotated database of patients with diverse forms of tuberculosis," PLOS ONE, Public Library of Science, vol. 15(1), pages 1-19, January.
    4. Matthias Schmid & Thomas Hielscher & Thomas Augustin & Olaf Gefeller, 2011. "A Robust Alternative to the Schemper–Henderson Estimator of Prediction Error," Biometrics, The International Biometric Society, vol. 67(2), pages 524-535, June.
    5. Paul Blanche & Jean‐François Dartigues & Jérémie Riou, 2022. "A closed max‐t test for multiple comparisons of areas under the ROC curve," Biometrics, The International Biometric Society, vol. 78(1), pages 352-363, March.
    6. Dendramis, Y. & Tzavalis, E. & Varthalitis, P. & Athanasiou, E., 2020. "Predicting default risk under asymmetric binary link functions," International Journal of Forecasting, Elsevier, vol. 36(3), pages 1039-1056.
    7. Tianyu Han & Sven Nebelung & Federico Pedersoli & Markus Zimmermann & Maximilian Schulze-Hagen & Michael Ho & Christoph Haarburger & Fabian Kiessling & Christiane Kuhl & Volkmar Schulz & Daniel Truhn, 2021. "Advancing diagnostic performance and clinical usability of neural networks via adversarial training and dual batch normalization," Nature Communications, Nature, vol. 12(1), pages 1-11, December.
    8. Ruosha Li & Limin Peng, 2017. "Assessing quantile prediction with censored quantile regression models," Biometrics, The International Biometric Society, vol. 73(2), pages 517-528, June.
    9. A. Gregory DiRienzo, 2009. "Flexible Regression Model Selection for Survival Probabilities: With Application to AIDS," Biometrics, The International Biometric Society, vol. 65(4), pages 1194-1202, December.
    10. Oded Rotem & Tamar Schwartz & Ron Maor & Yishay Tauber & Maya Tsarfati Shapiro & Marcos Meseguer & Daniella Gilboa & Daniel S. Seidman & Assaf Zaritsky, 2024. "Visual interpretability of image-based classification models by generative latent space disentanglement applied to in vitro fertilization," Nature Communications, Nature, vol. 15(1), pages 1-19, December.
    11. Seung Seog Han & Ik Jun Moon & Seong Hwan Kim & Jung-Im Na & Myoung Shin Kim & Gyeong Hun Park & Ilwoo Park & Keewon Kim & Woohyung Lim & Ju Hee Lee & Sung Eun Chang, 2020. "Assessment of deep neural networks for the diagnosis of benign and malignant skin neoplasms in comparison with dermatologists: A retrospective validation study," PLOS Medicine, Public Library of Science, vol. 17(11), pages 1-21, November.
    12. Hajime Uno & Tianxi Cai & Lu Tian & L. J. Wei, 2011. "Graphical Procedures for Evaluating Overall and Subject-Specific Incremental Values from New Predictors with Censored Event Time Data," Biometrics, The International Biometric Society, vol. 67(4), pages 1389-1396, December.
    13. Tianxi Cai & Thomas A Gerds & Yingye Zheng & Jinbo Chen, 2011. "Robust Prediction of t-Year Survival with Data from Multiple Studies," Biometrics, The International Biometric Society, vol. 67(2), pages 436-444, June.
    14. Shu Jiang & Jiguo Cao & Bernard Rosner & Graham A. Colditz, 2023. "Supervised two‐dimensional functional principal component analysis with time‐to‐event outcomes and mammogram imaging data," Biometrics, The International Biometric Society, vol. 79(2), pages 1359-1369, June.
    15. Rebecca Payne & Ming Yang & Yingye Zheng & Majken K. Jensen & Tianxi Cai, 2016. "Robust risk prediction with biomarkers under two‐phase stratified cohort design," Biometrics, The International Biometric Society, vol. 72(4), pages 1037-1045, December.
    16. Schmid, Matthias & Tutz, Gerhard & Welchowski, Thomas, 2018. "Discrimination measures for discrete time-to-event predictions," Econometrics and Statistics, Elsevier, vol. 7(C), pages 153-164.
    17. Ruosha Li & Jing Ning & Ziding Feng, 2022. "Estimation and inference of predictive discrimination for survival outcome risk prediction models," Lifetime Data Analysis: An International Journal Devoted to Statistical Methods and Applications for Time-to-Event Data, Springer, vol. 28(2), pages 219-240, April.
    18. He Kevin & Zhou Xiang & Jiang Hui & Wen Xiaoquan & Li Yi, 2018. "False discovery control for penalized variable selections with high-dimensional covariates," Statistical Applications in Genetics and Molecular Biology, De Gruyter, vol. 17(6), pages 1-11, December.
    19. Susana Díaz-Coto & Pablo Martínez-Camblor & Sonia Pérez-Fernández, 2020. "smoothROCtime: an R package for time-dependent ROC curve estimation," Computational Statistics, Springer, vol. 35(3), pages 1231-1251, September.
    20. Shashank Shetty & Ananthanarayana V S. & Ajit Mahale, 2022. "MS-CheXNet: An Explainable and Lightweight Multi-Scale Dilated Network with Depthwise Separable Convolution for Prediction of Pulmonary Abnormalities in Chest Radiographs," Mathematics, MDPI, vol. 10(19), pages 1-29, October.

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:gam:jforec:v:6:y:2024:i:2:p:22-417:d:1402397. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: MDPI Indexing Manager (email available below). General contact details of provider: https://www.mdpi.com .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.