An efficient joint model for high dimensional longitudinal and survival data via generic association features
Biometrics · 2024
DOI: 10.1093/biomtc/ujae149 ↗Kurz gefasst
Fachartikel in der Zeitschrift Biometrics, 2024. Verfasst von Van Tuan Nguyen und 6 weiteren Autor:innen. Beteiligte Einrichtung: LOPF, Califrais' Machine Learning Lab, Paris F-75010, France. Hauptthemen laut MEDLINE: Algorithms, Models, Statistical, Computer Simulation, Monte Carlo Method. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 50 Literaturangaben. Gefördert von: French National Cancer Institute.
Abstract (Originalsprache)
This paper introduces a prognostic method called FLASH that addresses the problem of joint modeling of longitudinal data and censored durations when a large number of both longitudinal and time-independent features are available. In the literature, standard joint models are either of the shared random effect or joint latent class type. Combining ideas from both worlds and using appropriate regularization techniques, we define a new model with the ability to automatically identify significant prognostic longitudinal features in a high-dimensional context, which is of increasing importance in many areas such as personalized medicine or churn prediction. We develop an estimation methodology based on the expectation–maximization algorithm and provide an efficient implementation. The statistical performance of the method is demonstrated both in extensive Monte Carlo simulation studies and on publicly available medical datasets. Our method significantly outperforms the state-of-the-art joint models in terms of C-index in a so-called “real-time” prediction setting, with a computational speed that is orders of magnitude faster than competing methods. In addition, our model automatically identifies significant features that are relevant from a practical point of view, making it interpretable, which is of the greatest importance for a prognostic algorithm in healthcare.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- LOPF, Califrais' Machine Learning Lab, Paris F-75010, France.
- Laboratoire de Probabilités, Statistique et Modélisation, Paris F-75005, France.
- Univ. Bordeaux, INSERM U1219, Bordeaux F-33076, France.
- INSERM, Centre de recherche des Cordeliers, Univ. Sorbonne, Univ. Paris Cité, Paris F-75006, France.
- HeKA, Inria Paris, Paris F-75015, France.
- French National Rare Disease Registry (BNDMR), Greater Paris University Hospitals (AP-HP), Paris F-75610, France.
ORCID-Profile: Sarah Zohar, Anne-Sophie Jannot
Förderung
- French National Cancer Institute
Zitieren
APA
Nguyen, V. T., Fermanian, A., Barbieri, A., Zohar, S., Jannot, A. S., Bussy, S., & Guilloux, A. (2024). An efficient joint model for high dimensional longitudinal and survival data via generic association features. Biometrics. https://doi.org/10.1093/biomtc/ujae149
Vancouver
Nguyen VT, Fermanian A, Barbieri A, Zohar S, Jannot AS, Bussy S, et al. An efficient joint model for high dimensional longitudinal and survival data via generic association features. Biometrics. 2024. doi:10.1093/biomtc/ujae149
BibTeX
@article{Nguyen2024,
author = {Nguyen, Van Tuan and Fermanian, Adeline and Barbieri, Antoine and Zohar, Sarah and Jannot, Anne-Sophie and Bussy, Simon and Guilloux, Agathe},
title = {An efficient joint model for high dimensional longitudinal and survival data via generic association features},
journal = {Biometrics},
year = {2024},
doi = {10.1093/biomtc/ujae149}
}
EUVIMED