Acoustic Localisation of Coronary Artery Stenosis

在联合王国，2006年冠心病（CHD）造成的死亡人数占所有死亡人数的16%以上（571,034人中有94,381人）。其中，95%（89,817）发生在55岁以上的人群中。保健费用估计为32亿欧元（人均50欧元），工作日损失造成的额外经济费用估计为39亿欧元，对病人的非正式护理费用为18亿欧元。冠心病是一个昂贵的杀手。它给纳税人带来了巨大的负担，每年花费近90亿元，随着人口老龄化的公认和不可阻挡的趋势，这一负担将继续逐年增加。冠心病是指动脉粥样硬化疾病的一种，动脉粥样硬化斑块（脂肪和钙沉积）在动脉壁上积聚形成部分阻塞，从而导致心肌缺血（流向心肌的血液不足）。随着时间的推移，所谓的易损斑块会突然破裂，并激活人体的凝血机制。这阻塞了动脉，导致（最常见的）心肌梗死：“心脏病发作”。目前没有冠心病的“标准”筛查工具。咨询医生的病人已经有一些不适，随后的诊断需要高度专业化的医生的干预和检查。我们提出了一个概念验证的调查，将计算应用数学与生物技术联系起来。一个成功的结果将为冠心病提供一种相对便宜的筛查和诊断工具，可以针对“高危”人群。动脉狭窄有一个声学特征（杂音），这是由产生的湍流血流冲击动脉壁引起的。这导致低振幅位移波（横波）穿过胸部，然后在胸部表面表现为扰动。这些干扰可以通过在皮肤上放置传感器进行非侵入性测量。动脉壁波的产生、通过胸腔的传播以及在胸腔表面的出现，都可以用一个详细的数学模型来描述，这个模型描述了人体组织（心脏、肺、肌肉等）的粘弹性。整个模型可以在软件中作为虚拟胸腔进行模拟，从而避免了在早期概念验证开发阶段对真人进行临床试验的需要。我们建议在理论和软件上开发和实现这个虚拟胸腔，并通过实验来验证它，以便从对高危人群有效的“早期筛查过程”的角度来评估这种方法。这项任务包括两个部分（正解和反解），这两个部分都将在一个真实的胸部机械模型上进行校准和实验。具体来说，虚拟箱子将被数学表述，计算实现和实验测试。对动脉紊乱的初步猜测将通过直接问题预测胸壁的表面紊乱。这些值与实测值之间的差异将形成一个迭代反求解程序，该程序将修改动脉扰动，直到测量值与计算值之间的差异最小化。一旦这个直接反解迭代完成，表面测量就被数学和软件解码为动脉影响。这些可以潜在地用于指示狭窄的存在、大小、位置和形态。正是这种诊断的潜力，本项目试图通过一项基础研究进行调查和评估，包括数学建模和分析，计算模拟，并通过对充满组织模拟凝胶的胸部模型进行实验验证。

In the UK in 2006 coronary heart disease (CHD) caused over 16% of all deaths (94,381 out of 571,034). Of these, 95% (89,817) occurred in people over the age of fifty-five. The cost of health care was estimated at 3.2 billion (50 per capita), the additional economic cost due to lost working days has been estimated at 3.9 billion, and the cost of informal care of patients, at 1.8 billion.Coronary heart disease is an expensive killer. It places a huge burden on the taxpayer, costing nearly 9 billion per year which, with the acknowledged and inexorable trend toward an ageing population, will continue to grow year by year.The term CHD is one of a number that refer to the disease of atherosclerosis, wherein atheromatous plaques (fat and calcium deposits) accumulate in an artery wall to form a partial blockage and thereby cause myocardial ischaemia (inadequate blood flow to the heart muscle). In time, so-called vulnerable plaques undergo a sudden rupture and activate the body's blood-clotting mechanism. This occludes the artery and leads to (the most common form of) myocardial infarction: a 'heart attack'.There is currently no 'standard' screening tool for CHD. Patients who consult their doctor are already in some discomfort and the subsequent diagnosis requires the intervention of, and examination by, highly specialized medical practitioners.We propose a proof-of-concept investigation which connects computational applied mathematics to biotechnology. A successful outcome would provide a relatively cheap screening and diagnosis tool for CHD which could be targeted towards 'at-risk' population groups.An arterial stenosis has an acoustic signature (bruit) which is triggered by the resulting turbulent blood flow impacting on the artery walls. This causes low amplitude displacement waves (shear waves) to travel through the chest, which then manifest themselves as disturbances on the chest surface. These disturbances can be measured non-invasively by placing sensors on the skin.The generation of waves at the artery wall, their transmission through the chest, and their appearance at the chest surface, can all be described by a detailed mathematical model which describes the viscoelastic nature of human tissue (heart, lungs, muscle etc). The entire model can be simulated in software as a virtual chest thus obviating the need, in the early proof-of-concept development stage, for clinical tests on real people.We propose to develop and implement this virtual chest in theory and in software and to validate it by experiment in order to evaluate this approach in terms of an effective 'early days screening process' for an at-risk population. This task consists of two parts (the direct and inverse solver) both of which will be calibrated and tested by experiments on a realistic mechanical model of the chest.Specifically, the virtual chest will be formulated mathematically, implemented computationally and tested experimentally. An initial guess at the arterial disturbance will, via the direct problem, predict the surface disturbances at the chest wall. The difference between these and measured values will form an iterative inverse solver procedure which will modify the arterial disturbance until the difference between the measured and computed values is minimised.Once this direct-inverse solution iteration has finished, the surface measurements have been decoded by the mathematics and software into the arterial impacts. These can, potentially, be used to indicate the presence, size, location and morphology of the stenosis. It is this potential for diagnosis that this project seeks to investigate and evaluate through a fundamental study involving mathematical modelling and analysis, computational simulation, and validation through experimentation on chest phantoms filled with tissue-mimicking gel.