Uncertainty-Aware Longitudinal Forecasting of Alzheimer's Disease Progression Using Deep Learning

2026-06-23

AI Summary

This study addresses the challenge of reliably forecasting the longitudinal progression of Alzheimer’s disease using deep learning, where existing methods struggle with multi-step prediction accuracy and fail to provide fine-grained characterization of predictive uncertainty. To overcome these limitations, the authors propose a probabilistic deep learning framework that integrates an enhanced Temporal Fusion Transformer encoder, a CORAL-based ordinal output layer, and an autoregressive mixture density network, augmented with Bootstrap ensembling. This approach uniquely disentangles and quantifies aleatoric and epistemic uncertainties in longitudinal predictions. Evaluated on the ADNI dataset, the method significantly outperforms baseline models—particularly in discriminating between mild cognitive impairment and dementia—and generates five-year trajectories of clinical diagnoses and biomarkers with well-calibrated uncertainty bounds, achieving approximately 90% prediction interval coverage. Strong generalization performance is further demonstrated on the OASIS-3 dataset.

Abstract

Longitudinal modelling of Alzheimer's disease progression is clinically useful only if it can describe not just the most likely next diagnosis, but how a patient may evolve over time and how reliable that forecast is. Most deep learning approaches reduce this problem to single-step classification, treating cognitively normal, mild cognitive impairment, and dementia as flat categories while providing limited insight into how uncertainty accumulates across future visits. We propose a probabilistic framework that combines ordinal diagnosis prediction, multi-horizon trajectory generation, and decomposed uncertainty estimation. A Temporal Fusion Transformer encoder is adapted with a CORAL ordinal output layer, asymmetric loss weighting, and converter oversampling to respect disease-stage ordering and improve sensitivity to MCI-to-dementia transitions. Conditioned on the learned patient-context representation, an autoregressive Mixture Density Network generates five-year probabilistic trajectories for diagnosis state, CDR Sum of Boxes, MMSE orientation, and hippocampal volume. On ADNI, the model outperforms linear, recurrent, and transformer baselines for next-visit diagnosis prediction, with the strongest gains on MCI-versus-dementia discrimination. Generated trajectories achieve near-nominal 90% credible interval coverage, widening uncertainty across the forecast horizon, and biomarker dynamics consistent with expected Alzheimer's disease progression. We further separate aleatoric from epistemic uncertainty using analytic mixture variance and a five-member bootstrap ensemble, which provides the strongest encoder diversity and output-level epistemic signal. Epistemic uncertainty is higher for rare progression archetypes, MCI and dementia patients, and under external evaluation on OASIS-3, where it increases alongside prediction error.

Problem

• Alzheimer's disease
• longitudinal forecasting
• uncertainty quantification
• disease progression
• deep learning

Innovation

• ordinal forecasting
• uncertainty decomposition
• Temporal Fusion Transformer
• Mixture Density Network
• Alzheimer's disease progression