Uncertainty-Calibrated Prediction of Randomly-Timed Biomarker Trajectories with Conformal Bands

2025-11-17

AI Summary

Uncertainty calibration for predicting brain biomarker trajectories under irregular sampling in Alzheimer’s disease remains challenging. Method: This paper proposes a group-conditional conformal prediction framework tailored to population-level heterogeneity. We design a novel nonconformity score specifically suited for sparse and asynchronous longitudinal data, and integrate multiple state-of-the-art time-series models—including LSTM and TCN—to generate individually calibrated prediction intervals. A group-conditional calibration mechanism ensures rigorous coverage guarantees on real-world neuroimaging data. Contributions/Results: Experiments demonstrate tighter prediction intervals with coverage error below 1.2%. Compared to conventional risk scoring, our method improves high-risk individual identification by 17.5%, significantly enhancing the safety and reliability of clinical decision-making.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Relational Probabilistic Models
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

Despite recent progress in predicting biomarker trajectories from real clinical data, uncertainty in the predictions poses high-stakes risks (e.g., misdiagnosis) that limit their clinical deployment. To enable safe and reliable use of such predictions in healthcare, we introduce a conformal method for uncertainty-calibrated prediction of biomarker trajectories resulting from randomly-timed clinical visits of patients. Our approach extends conformal prediction to the setting of randomly-timed trajectories via a novel nonconformity score that produces prediction bands guaranteed to cover the unknown biomarker trajectories with a user-prescribed probability. We apply our method across a wide range of standard and state-of-the-art predictors for two well-established brain biomarkers of Alzheimer's disease, using neuroimaging data from real clinical studies. We observe that our conformal prediction bands consistently achieve the desired coverage, while also being tighter than baseline prediction bands. To further account for population heterogeneity, we develop group-conditional conformal bands and test their coverage guarantees across various demographic and clinically relevant subpopulations. Moreover, we demonstrate the clinical utility of our conformal bands in identifying subjects at high risk of progression to Alzheimer's disease. Specifically, we introduce an uncertainty-calibrated risk score that enables the identification of 17.5% more high-risk subjects compared to standard risk scores, highlighting the value of uncertainty calibration in real-world clinical decision making. Our code is available at github.com/vatass/ConformalBiomarkerTrajectories.

Problem

• Predicting biomarker trajectories with guaranteed uncertainty calibration for clinical safety
• Extending conformal prediction to handle randomly-timed patient visit patterns
• Developing risk scores that identify more high-risk Alzheimer's disease patients

Innovation

• Conformal prediction for biomarker trajectory uncertainty calibration
• Novel nonconformity score guarantees user-prescribed coverage probability
• Group-conditional conformal bands address population heterogeneity