Symptom-Driven Personalized Proton Pump Inhibitors Therapy Using Bayesian Neural Networks and Model Predictive Control

2025-07-13

AI Summary

Long-term, high-dose proton pump inhibitor (PPI) therapy for acid-related disorders carries significant clinical risks, while invasive intragastric pH monitoring is unsustainable and interpatient pharmacodynamic variability hampers precise acid suppression. To address this, we propose the first non-invasive, personalized PPI dosing framework: it integrates patient-reported reflux and dyspeptic symptoms with a Bayesian neural network—enabling both symptom prediction and predictive uncertainty quantification—and an opportunity-constrained model predictive control (MPC) scheme to dynamically optimize dosing regimens. Crucially, this approach guarantees robust acid suppression without requiring real-time pH monitoring. In silico validation demonstrates that, compared to fixed-dose regimens, our method reduces total PPI exposure by 65% while maintaining acid suppression success rates above 95%, thereby substantially improving therapeutic safety and personalization.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Sampling/Simulation-based Search
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Proton Pump Inhibitors (PPIs) are the standard of care for gastric acid disorders but carry significant risks when administered chronically at high doses. Precise long-term control of gastric acidity is challenged by the impracticality of invasive gastric acid monitoring beyond 72 hours and wide inter-patient variability. We propose a noninvasive, symptom-based framework that tailors PPI dosing solely on patient-reported reflux and digestive symptom patterns. A Bayesian Neural Network prediction model learns to predict patient symptoms and quantifies its uncertainty from historical symptom scores, meal, and PPIs intake data. These probabilistic forecasts feed a chance-constrained Model Predictive Control (MPC) algorithm that dynamically computes future PPI doses to minimize drug usage while enforcing acid suppression with high confidence - without any direct acid measurement. In silico studies over diverse dietary schedules and virtual patient profiles demonstrate that our learning-augmented MPC reduces total PPI consumption by 65 percent compared to standard fixed regimens, while maintaining acid suppression with at least 95 percent probability. The proposed approach offers a practical path to personalized PPI therapy, minimizing treatment burden and overdose risk without invasive sensors.

Problem

• Noninvasive PPI dosing based on symptom patterns
• Minimizing drug usage while ensuring acid suppression
• Reducing PPI consumption compared to fixed regimens

Innovation

• Bayesian Neural Network predicts symptom patterns
• Model Predictive Control optimizes PPI dosing
• Noninvasive framework reduces drug usage significantly