Toward an accuracy-efficiency balanced model for modeling high intensity focused ultrasound

摘要 在过去的二十年里，高强度聚焦超声(HIFU)被大量研究用于治疗 各种疾病和医疗条件。作为一种可以触及深层组织的非侵入性手术方式， HIFU有可能给治疗带来革命性的变化。一种通用、快速而又足够精确的超声数值计算 模型对HIFU至关重要。这样的数值模式可以作为深入研究的强大平台 HIFU和作为HIFU技术临床翻译的跳板。一些特定的应用包括测试 新的超声治疗方案，了解某些HIFU技术的机制，以及快速治疗 计划。 存在两种常见的HIFU技术：基于热的HIFU技术和基于冲击波的HIFU技术 技术。而第一种类型使用相对较低的压力连续波(CW)，并且主要引起热 凝固，第二种类型使用高压冲击波脉冲来引起组织的机械分割。 到目前为止，还没有一个模型能够有效和准确地在三维(3D)中模拟这两种类型的HIFU 大规模、复杂、异质的生物组织。 在拟议的研究中，我们将解决HIFU社区对新颖、准确- 效率平衡的数值模型。在目标1中，我们将创造一个全面、准确和效率平衡的 用于HIFU建模的算法。通过在Westvelt方程中考虑组织的非均质性，一个修正的波-- 任意介质中线性/非线性波场的矢量频域(M-WVFD)法 我们将首次对非均质介质进行系统研究。最终的模型预计将是 至少比最先进的精确模型快两个数量级，而且仍然具有很高的精度。 在目标2中，将通过研究各种HIFU对M-WVFD方法进行数值和实验验证 有问题。为了研究波，这些实验将包括体模和体外的人类头骨。 弱非均质和强非均质介质中的传播。目标3应专注于软件工程，以便 开发的算法可以集成到开放源码软件包中，并可以被 HIFU社区。

Abstract High intensity focused ultrasound (HIFU) has been heavily investigated over the past two decades for treating a wide range of diseases and medical conditions. As a non-invasive surgical modality that can reach deep tissues, HIFU has the potential to revolutionize therapy. A versatile and fast yet sufficiently accurate ultrasound numerical model is vital for HIFU. Such a numerical model could serve as a powerful platform for in-depth investigation of HIFU and as a springboard for clinical translation of HIFU techniques. Some specific applications include testing new sonication protocols, understanding the mechanism of certain HIFU techniques, and rapid treatment planning. Two common HIFU techniques exist: they are the thermal based HIFU technique and shock wave based HIFU technique. While the first type uses relatively low pressure continuous wave (CW) and primarily causes thermal coagulation, the second type uses high pressure shock wave pulses to cause mechanical fractionation of tissue. To this date, no model can efficiently and accurately model these two types of HIFU in three-dimensional (3D) large-scale, complex, heterogeneous biological tissue. In the proposed research, we shall address a longstanding need in the HIFU community for a novel, accuracy- efficiency balanced numerical model. In Aim 1, we shall create a versatile, accuracy-efficiency balanced algorithm for HIFU modeling. By considering tissue heterogeneities in the Westervelt equation, a modified wave- vector-frequency-domain (M-WVFD) method for predicting linear/nonlinear wave fields in arbitrarily heterogeneous media will be systematically investigated for the first time. The resulting model is expected to be at least two orders of magnitude faster than the state-of-the-art “accurate” models and still have high accuracy. In Aim 2, the M-WVFD method will be numerically and experimentally validated by investigating a variety of HIFU problems. The experiments will involve both phantom and ex-vivo human skulls in order to investigate wave propagation in weakly and strongly heterogeneous media. Aim 3 shall focus on software engineering, so that the developed algorithms can be integrated into an open-source software package and can be widely adopted by the HIFU community.