Vicini P, Bonadonna R C, Lehtovirta M, Groop L C, Cobelli C
Department of Electronics and Informatics, University of Padova, Italy.
Ann Biomed Eng. 1998 Sep-Oct;26(5):764-74. doi: 10.1114/1.64.
Distributed models of blood-tissue exchange are widely used to measure kinetic events of various solutes from multiple tracer dilution experiments. Their use requires, however, a careful description of blood flow heterogeneity along the capillary bed. Since they have mostly been applied in animal studies, direct measurement of the heterogeneity distribution was possible, e.g., with the invasive microsphere method. Here we apply distributed modeling to a dual tracer experiment in humans, performed using an intravascular (indocyanine green dye, subject to distribution along the vascular tree and confined to the capillary bed) and an extracellular ([3H]-D-mannitol, tracing passive transcapillary transfer across the capillary membrane in the interstitial fluid) tracer. The goal is to measure relevant parameters of transcapillary exchange in human skeletal muscle. We show that assuming an accurate description of blood flow heterogeneity is crucial for modeling, and in particular that assuming for skeletal muscle the well-studied cardiac muscle blood flow heterogeneity is inappropriate. The same reason prevents the use of the common method of estimating the input function of the distributed model via deconvolution, which assumes a known blood flow heterogeneity, either defined from literature or measured, when possible. We present a novel approach for the estimation of blood flow heterogeneity in each individual from the intravascular tracer data. When this newly estimated blood flow heterogeneity is used, a more satisfactory model fit is obtained and it is possible to reliably measure parameters of capillary membrane permeability-surface product and interstitial fluid volume describing transcapillary transfer in vivo.
分布式血液-组织交换模型被广泛用于通过多次示踪剂稀释实验来测量各种溶质的动力学事件。然而，其应用需要仔细描述沿毛细血管床的血流异质性。由于它们大多应用于动物研究，因此可以直接测量异质性分布，例如采用侵入性微球法。在此，我们将分布式建模应用于人体的双示踪剂实验，该实验使用了一种血管内示踪剂（吲哚菁绿染料，沿血管树分布并局限于毛细血管床）和一种细胞外示踪剂（[3H]-D-甘露醇，追踪间质液中跨毛细血管膜的被动跨毛细血管转运）。目的是测量人体骨骼肌中跨毛细血管交换的相关参数。我们表明，假设对血流异质性有准确描述对于建模至关重要，特别是对于骨骼肌而言，假设采用已充分研究的心肌血流异质性是不合适的。同样的原因也使得无法使用通过去卷积估计分布式模型输入函数的常用方法，该方法假设已知血流异质性，要么是从文献中定义的，要么是在可能的情况下测量得到的。我们提出了一种从血管内示踪剂数据估计每个个体血流异质性的新方法。当使用这种新估计的血流异质性时，可以获得更令人满意的模型拟合，并且能够可靠地测量描述体内跨毛细血管转运的毛细血管膜通透率-表面积乘积和间质液体积的参数。
