Bracamonte, J. H. | Search bioRxiv

bioRxiv2026

Population-level emulation enables fast Bayesian calibration of cardiac digital twins

Purpose: Patient-specific cardiac digital twins have the potential to guide individualized therapy, but generating these in a routine clinical workflow remains challenging. This is especially true when the twins must carry quantified uncertainty: Bayesian calibration requires thousands of simulations per patient, and the emulators that reduce this cost are typically trained one patient at a time. We aimed to train a single emulator across a patient cohort, so that a new patient can be calibrated in minutes without new simulations. Methods: We trained a population-level Gaussian process emulator whose inputs included a 12-dimensional latent encoding of the clinical features of each patient, so that a single emulator represents the whole cohort. We compared spending simulation budgets on resampling the real cohort with augmenting the cohort with digital chimeras, synthetic patients interpolated between real patients. Seven parameters of a closed-loop heart and circulation model were calibrated by Markov chain Monte Carlo to ten clinical biomarkers in 26 patients with left bundle branch block, each using an emulator trained without that patient, and the calibrated posteriors were compared against the full simulator. Results: A single emulator, trained without the held-out patient, calibrated the digital twin of each patient in 59.7-63.0 s per patient, versus 15 h with the simulator. Resampling the real cohort was the more efficient training strategy, bringing every output to R2 greater than or equal to 0.9, which augmentation did not achieve for the same simulation budget. For held-out patients, both strategies performed comparably (median leave-one-out normalized root-mean-square error 11.0% vs. 10.7%; p = 0.745), and error rose with distance from the cohort centroid (Spearman rho = +0.406). The calibrated models showed no cohort-wide bias (median standardized residual -0.12 against the clinical targets), and 95% predictive intervals contained 78.5% of targets, with the largest misfit in systolic blood pressure. Conclusion: Removing the need for new simulations for every new patient is a meaningful step toward uncertainty-aware cardiac digital twins that fit within a routine clinical workflow. Because the simulator, surrogate and sampler are interchangeable, the approach used here can be applied to other (cardiac) models and cohorts.

bioengineering↗

bioRxiv2024

Contributions of mechanical loading and hormonal changes to eccentric hypertrophy during volume overload: a Bayesian analysis using logic-based network models.

Primary mitral regurgitation (MR) is a pathology that alters mechanical loading on the left ventricle, triggers an array of compensatory neurohormonal responses, and induces a distinctive ventricular remodeling response known as eccentric hypertrophy. Drug therapies may alleviate symptoms, but only mitral valve repair or replacement can provide significant recovery of cardiac function and dimensions. Questions remain about the optimal timing of surgery, with 20% of patients developing systolic dysfunction post-operatively despite being treated according to the current guidelines. Thus, better understanding of the hypertrophic process in the setting of ventricular volume overload (VO) is needed to improve and better personalize the management of MR. To address this knowledge gap, we employ a Bayesian approach to combine data from 70 studies on experimental volume overload in dogs and rats and use it to calibrate a logic-based network model of hypertrophic signaling in myocytes. The calibrated model predicts that growth in experimental VO is mostly driven by the neurohormonal response, with an initial increase in myocardial tissue stretch being compensated by subsequent remodeling fairly early in the time course of VO. This observation contrasts with a common perception that volume-overload hypertrophy is driven primarily by increased myocyte strain. The model reproduces many aspects of 43 studies not used in its calibration, including infusion of individual hypertrophic agonists alone or in combination with various drugs commonly employed to treat heart failure, as well as administration of some of those drugs in the setting of experimental volume overload. We believe this represents a promising approach to using the known structure of an intracellular signaling network to integrate information from multiple studies into quantitative predictions of the range of expected responses to potential interventions in the complex setting of cardiac hypertrophy driven by a combination of hormonal and mechanical factors. Author SummaryMitral valve regurgitation is a common heart disease in which a malfunctioning valve allows part of the blood pumped by the heart to flow in the wrong direction. This condition overloads the heart by making it pump more blood volume than normal; the heart temporarily adapts by growing in mass and volume, but if untreated the condition can ultimately lead to heart failure and death. The most effective treatment is to surgically repair the valve; however, in many patients heart function deteriorates even after a successful surgery. Many researchers have studied this condition by experimentally overloading the hearts of dogs and rats, producing large amounts of data on the resulting geometric, mechanical, and biologic changes. Yet it has been difficult to translate those studies into effective selection and timing of treatment in all patients. In this work we integrate experimental data reported from 70 research articles on experimental volume overload through a simple model of heart mechanics and a more complex model of the molecular signaling pathways inside heart cells. We use a statistical approach to calibrate the computational model, so that it can predict not only average responses but also the degree of expected uncertainty for each prediction. We then use the model to explore how the heart responds to different combinations of drugs and hormones, and potential treatments during overload.

bioengineering↗