A Probabilistic Neural Twin for Treatment Planning in Peripheral Pulmonary Artery Stenosis

John D. Lee,Jakob Richter,Martin R. Pfaller,Jason M. Szafron,Karthik Menon,Andrea Zanoni,Michael R. Ma,Jeffrey A. Feinstein,Jacqueline Kreutzer,Alison L. Marsden,Daniele E. Schiavazzi

2023-12-01

Abstract:The substantial computational cost of high-fidelity models in numerical hemodynamics has, so far, relegated their use mainly to offline treatment planning. New breakthroughs in data-driven architectures and optimization techniques for fast surrogate modeling provide an exciting opportunity to overcome these limitations, enabling the use of such technology for time-critical decisions. We discuss an application to the repair of multiple stenosis in peripheral pulmonary artery disease through either transcatheter pulmonary artery rehabilitation or surgery, where it is of interest to achieve desired pressures and flows at specific locations in the pulmonary artery tree, while minimizing the risk for the patient. Since different degrees of success can be achieved in practice during treatment, we formulate the problem in probability, and solve it through a sample-based approach. We propose a new offline-online pipeline for probabilsitic real-time treatment planning which combines offline assimilation of boundary conditions, model reduction, and training dataset generation with online estimation of marginal probabilities, possibly conditioned on the degree of augmentation observed in already repaired lesions. Moreover, we propose a new approach for the parametrization of arbitrarily shaped vascular repairs through iterative corrections of a zero-dimensional approximant. We demonstrate this pipeline for a diseased model of the pulmonary artery tree available through the Vascular Model Repository.

Medical Physics,Artificial Intelligence,Machine Learning,Numerical Analysis,Computation

What problem does this paper attempt to address?

This paper discusses how to use the probabilistic neural network twin model for treatment planning in the treatment of peripheral pulmonary artery stenosis (PPAS). Currently, high-precision numerical hemodynamic models are mainly used for offline treatment planning due to the high computational cost. The researchers propose a new offline-online pipeline, which combines offline boundary condition assimilation, model dimensionality reduction and training data generation, as well as online marginal probability estimation, to achieve probabilistic real-time treatment planning. This method takes into account the probability of different treatment outcomes and solves this problem through a sample-based approach. The paper introduces an application for repairing multiple stenoses in PPAS, which can be treated through endovascular intervention or surgery, with the goal of achieving desired pressure and flow at specific locations while minimizing patient risk. A new vascular repair parameterization method is proposed in the study to handle arbitrary shapes of vascular repairs through iterative correction of zero-dimensional approximations. In the study, the authors constructed a model of the disease's pulmonary arterial tree and simulated it using zero-dimensional and three-dimensional models. They also designed a neural network twin model that trains a fully connected neural architecture by generating a dataset to predict the joint probability distribution of treatment outcomes. This method can handle uncertainty in treatment and provides support for clinical decision-making. The structure of the paper includes a discussion of the probabilistic framework, a description of patient-specific pulmonary anatomy and hemodynamic models, an introduction to the zero-dimensional model, the construction and validation of the neural network, and examples of clinical problem solving. Finally, the paper discusses the potential impact of this method on future virtual treatment planning tools.

A Probabilistic Neural Twin for Treatment Planning in Peripheral Pulmonary Artery Stenosis

A probabilistic neural twin for treatment planning in peripheral pulmonary artery stenosis

Rapid Prediction Method for Nonlinear Expansion Process of Medical Vascular Stent

Computational replication of the patient-specific stenting procedure for coronary artery bifurcations: From OCT and CT imaging to structural and hemodynamics analyses

Physics driven reduced order model for real time blood flow simulations

AI-powered multimodal modeling of personalized hemodynamics in aortic stenosis

Physics-informed neural networks (PINNs) for 4D hemodynamics prediction: An investigation of optimal framework based on vascular morphology

Virtual Planning and Patient-Specific Graft Design for Aortic Repairs

Hemodynamic-based virtual surgery design of double-patch repair for pulmonary arterioplasty in Tetralogy of Fallot

Computational Pre-surgical Planning of Arterial Patch Reconstruction: Parametric Limits and In Vitro Validation

Numerical modeling of hemodynamics scenarios of patient-specific coronary artery bypass grafts

226 Advancing Neurovascular Diagnostics for Abnormal Hemodynamic Conditions Through AI-Driven Physics-informed Neural Networks

Deforming Patient-Specific Models of Vascular Anatomies to Represent Stent Implantation via Extended Position Based Dynamics

Uncertainty quantification in virtual surgery hemodynamics predictions for single ventricle palliation

Patient-specific coronary angioplasty simulations -- a mixed-dimensional finite element modeling approach

In silico modeling for personalized stenting in aortic coarctation

Constitutive neural networks for main pulmonary arteries: Discovering the undiscovered

A physics-informed deep learning framework for modeling of coronary in-stent restenosis

Numerical modelling of a peripheral arterial stenosis using dimensionally reduced models and kernel methods

Software for optimized virtual stenting of patient-specific coronary arteries reconstructed from angiography images

Patient-specific computational simulation of coronary artery bypass grafting