Our platform integrates widely used statistical models for analyzing time-to-event data.
Below is a brief overview of the models currently implemented and others that can be extended in future developments:
Cox Proportional Hazards Model: A semi-parametric model that estimates hazard ratios while making no assumptions about the baseline hazard. It is the most widely used model in clinical research.
Weibull Model: A parametric survival model that can capture increasing, decreasing, or constant hazard rates, making it highly flexible for diverse clinical data.
Log-Normal Model: Assumes survival times follow a log-normal distribution, useful when hazard rates initially rise and then decline.
Additional Models of Interest
Exponential Model: The simplest survival model with a constant hazard over time, often used as a baseline.
Gompertz Model: Useful for modeling monotonically increasing hazard rates, frequently applied in demographic and aging studies.
Accelerated Failure Time (AFT) Models: A class of parametric models (e.g., log-logistic, generalized gamma) that directly model survival time instead of hazard.
Random Survival Forests: A non-parametric, machine-learning approach that handles non-linear effects and complex interactions between variables.
By combining classical statistical models with modern machine learning methods, our platform enables robust and interpretable survival analyses for real-world clinical datasets.
An AI-generated video by NotebookLM showcasing the core ideas behind the ClinicalStatAI theory — where automation meets intelligent research innovation.
A short video that offers an overview of ClinicalStatAI, our intelligent platform for automated research and analysis.