Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Non-linear survival analysis using neural networks.

Ruth M Ripley1, Adrian L Harris, Lionel Tarassenko

  • 1Department of Statistics, University of Oxford, Oxford OX1 3TG, UK. ruth@stats.ox.ac.uk

Statistics in Medicine
|February 26, 2004
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Generating exosome subtypes: diverse membrane origins and mergers.

Trends in cell biology·2026
Same author

SET1B Drives Sustained HIF activity and Disease Progression in Clear Cell Renal Cell Carcinoma.

Cancer research·2026
Same author

Editorial Expression of Concern: Direct targeting of Sec23a by miR-200s influences cancer cell secretome and promotes metastatic colonization.

Nature medicine·2026
Same author

A systems-based approach to uterine fibroids identifies differential splicing associated with abnormal uterine bleeding.

Communications medicine·2025
Same author

Tumor-specific CD8 T cell characterization in HR<sup>+</sup> breast cancer reveals an impaired antitumoral response in patients with lymph node metastasis.

Cell reports. Medicine·2025
Same author

Amyloid-β disrupts APP-regulated protein aggregation and dissociation from recycling endosomal membranes.

The EMBO journal·2025
Same journal

Interpretable Bayesian Modeling for Multireader Multicase Studies: Addressing Overdispersion and Limited Sample Size in Diagnostic Enhancement Evaluation.

Statistics in medicine·2026
Same journal

Adaptive Sequential Multiple Hypotheses Testing for Concomitant Vaccine Safety Surveillance.

Statistics in medicine·2026
Same journal

Novel Distance Regression for Repeated Outcomes With Missing Data: Applications to Longitudinal and Crossover Studies of Microbiome Beta-Diversity.

Statistics in medicine·2026
Same journal

Optimal Weighted Tests for Replication Studies and the 'Two-Trials Rule' With Multiple Hypotheses.

Statistics in medicine·2026
Same journal

Identifiable Copula-Double-Cox Models: A Fully Parametric Framework for Dependent Right-Censored Survival Data.

Statistics in medicine·2026
Same journal

Moving From Individualized Risk-Based Prevention to Benefit-Based Prevention: Estimating Individualized Life-Years Gained From Prevention Services as a Basis for Eligibility.

Statistics in medicine·2026
See all related articles

New neural network models enhance survival analysis by relaxing traditional regression assumptions. These flexible models implicitly fit non-linear predictors and time-varying covariate effects, improving predictions for diseases like breast cancer.

Area of Science:

  • Biostatistics
  • Machine Learning
  • Computational Biology

Background:

  • Traditional regression models in survival analysis rely on restrictive assumptions.
  • There is a need for more flexible models that can capture complex relationships between covariates and survival outcomes.
  • Neural networks offer a powerful framework for developing such advanced statistical models.

Purpose of the Study:

  • To introduce and evaluate novel survival analysis models based on multi-layer perceptrons (neural networks).
  • To demonstrate the ability of these models to relax traditional assumptions and incorporate non-linearities and time-varying effects.
  • To compare the predictive performance of these neural network models against traditional linear models.

Main Methods:

  • Development of seven distinct neural network survival models.

Related Experiment Videos

  • Utilizing cross-validation to determine the beneficial inclusion of model flexibility.
  • Comparing predictive accuracy with linear regression models for survival data.
  • Estimating the hazard function using the developed neural network models.

    • Main Results:

    • Neural network models successfully relaxed traditional regression assumptions, encompassing them as special cases.
    • Models demonstrated implicit fitting of non-linear predictors and time-varying covariate effects.
    • Cross-validation guided the incorporation of flexibility, enhancing model performance.
    • The models showed competitive or superior predictive accuracy compared to linear models in predicting breast cancer relapse time.

    Conclusions:

    • Neural network-based survival models offer a flexible and powerful alternative to traditional methods.
    • These models can improve the prediction of survival outcomes, such as time to relapse in breast cancer.
    • The approach facilitates the discovery of important regressors and allows for diverse hazard function modeling.