STTR Phase I: Feasibility Study of a Reduced Order Model for Calculating Fractional Flow Reserve (FFR) Using Angiographic Data

小企业技术转让（STTR）第一阶段项目的更广泛影响是改善冠状动脉疾病（CAD）患者的临床护理，并降低相关成本。CAD仍然是最常见的心脏病形式，困扰着超过1800万成年人，每年花费美国医疗保健系统超过900亿美元。改善治疗决策的诊断学进展落后于治疗的进展。拟议的研究将在人工智能（AI）软件平台内探索新的计算建模方法，以改善诊断和优化患者特定的治疗决策。价值主张是改善临床结果、降低医疗成本和挽救生命。该软件将提供竞争优势，作为一种更加用户友好和非侵入性的诊断方法，能够比现有的替代品更快，更准确地对CAD进行临床评估。该小型企业技术转让（STTR）第一阶段项目将推进冠状动脉疾病（CAD）的诊断，以优化患者特定的治疗决策。CAD患者通常经历血管造影术，由此使冠状动脉可视化以估计狭窄严重性并做出关于是否执行血管重建术的主观治疗决定（例如，支架植入术或旁路移植手术）。最近，血流储备分数（FFR），一种测量血管狭窄处压力梯度的方法，已证明在指导治疗决策时可改善结局。然而，考虑到可用介入器械的缺点和成本，FFR的采用仍然有限。最近，商业努力已经开发了基于计算机的方法来无创地估计FFR，尽管在FFR的边界值中具有有限的准确性。拟议的研究将使用新型机器学习和计算方法开发和校准降阶模型（ROM），以比现有解决方案更快、更准确、更集成的方式提供FFR测量。ROM将依靠图论的最新进展来增强血流的一维非线性公式。该方法将使用由地面实况3D Navier-Stokes解决方案生成的合成数据进行校准，并根据20名患者的FFR临床测量结果进行验证。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact of this Small Business Technology Transfer (STTR) Phase I project is to improve the clinical care of patients suffering from coronary artery disease (CAD) and to reduce the associated costs. CAD remains the most common form of heart disease, afflicting more than 18 million adults and costing the U.S. healthcare system over $90 billion annually. Advances in diagnostics to improve treatment decisions have lagged behind advances in therapies. The proposed research will explore novel computational modeling methods within an artificial intelligence (AI) software platform to improve diagnosis and optimize patient-specific treatment decisions. Value propositions are to improve clinical outcomes, reduce healthcare costs, and save lives. The software will provide competitive advantages as a more user-friendly and non-invasive diagnostic method capable of faster and more accurate clinical assessment of CAD than existing alternatives. This Small Business Technology Transfer (STTR) Phase I project will advance the diagnosis of coronary artery disease (CAD) to optimize patient-specific treatment decisions. CAD patients typically undergo an angiography procedure whereby coronary arteries are visualized to estimate stenosis severity and make subjective treatment decisions on whether to perform revascularization procedures (e.g., stenting or bypass graft surgeries). More recently Fractional Flow Reserve (FFR), a measure of the pressure gradient across the vessel stenosis, has demonstrated improved outcomes when guiding treatment decisions. However, adoption of FFR remains modest given shortcomings of available interventional devices and cost. Recently, commercial efforts have developed computer-based methods to estimate FFR non-invasively, although with limited accuracy in borderline values of FFR. The proposed research will develop and calibrate a Reduced Order Model (ROM) using novel machine learning and computational methods to provide FFR measures in a faster, more accurate, and more integrated manner with clinical workflows than existing solutions. The ROM will rely on recent advances in Graph Theory to enhance a 1D nonlinear formulation of blood flow. The method will be calibrated using synthetic data generated with ground truth 3D Navier-Stokes solutions and validated against clinical measurements of FFR from a cohort of 20 patients.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.