Quantifying Uncertainty in Alzheimer’s Disease Progression Modelling: A Variational Disease Progression Score Framework
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Predicting the course of Alzheimer’s disease for individual patients remains a major challenge due to the heterogeneity of disease expression and the sparsity of longitudinal data. We introduce a variational Disease Progression Score (DPS) framework that maps multimodal biomarker dynamics (Cerebrospinal fluid, neuroimaging, and cognitive assessments) onto a continuous latent timeline with quantified uncertainty. The framework combines a neural encoder, which infers subject-specific progression parameters from demographic and clinical features, with a cascade of logistic functions structured according to the amyloid cascade hypothesis. Applied to the Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort, the inferred timeline separated diagnostic groups it never observed (AUC 0.98 for cognitively normal vs Alzheimer’s Disease), and the estimated cascade strengths and biomarker orderings were consistent with the established sequence of Alzheimer’s pathology. The model produces individualised prognoses for previously unseen subjects from baseline data alone, with 95% credible intervals achieving 89-98% empirical coverage across biomarkers, and these predictions can be dynamically refined as new observations become available. The framework thus provides a biologically interpretable, uncertainty-aware index of disease severity, offering a probabilistic foundation for patient-level prognosis and precision monitoring in Alzheimer’s disease.
Author summary
Alzheimer’s disease develops gradually, and its course varies widely from person to person, which makes it hard to predict how any individual will progress. Accurate, individual-level predictions would help doctors intervene earlier and design better clinical trials, but the data available in practice are sparse and collected at irregular times, and most existing prediction tools offer little sense of how confident they are.
We developed a method that places each person on a shared disease timeline estimated from routinely collected information such as age, genetics, and a brief cognitive assessment. Moreover, our model reports the uncertainty of each prediction and refines both the prediction and its uncertainty as new measurements become available for a patient.
Using the ADNI dataset, we found that the model realigned patients according to their clinical diagnoses. It also recovered biological findings consistent with the established understanding of the disease. Its confidence estimates were reliable for most measurements. We hope this approach is a step forward in the development of prediction tools that clinicians can genuinely trust for individual patients.
