磁共振温度成像与生物传热模拟协同提升的探索研究及其在椎体肿瘤微波消融手术实时监测中的应用

Thermal ablation is a new and efficient treatment of tumor. But for application in spinal bone tumor which is in the vicinity of vulnerable spinal cord and blood vessels, accurate real time temperature monitoring is an important factor for success of safe and effective tumor thermal ablation. .In this project, a novel method is proposed to synergy MR thermometry (MRTI) and bio-heat transfer (BHT) model such that the combined two perform better than the one alone is able to do. The proposed method combines the forward problem and the inverse problem of BHT with MRTI. First, the structure image of MRI is utilized to establish the BHT model and calculate the temperature field of all regions as a forward problem. Second, the BHT model is introduced to cost function of temperature estimation in MRTI to further regularize the estimation and improve the accuracy and convergence rate. Then, the estimated temperature of MRTI is used as a feedback to correct the parameters in BHT as an inverse problem and help calculate the water-free areas more accurately. The modified BHT model is again applied to next MRTI process. By repeating this process iteratively, the BHT model and MRTI can be both improved while the convergence rate is increased, which can finally provide a reliable and fast temperature monitoring method for the surgeons.

热疗是治疗肿瘤的一种新的有效手段，而对于椎体骨肿瘤这一特殊应用，由于其附近有易受损伤的脊髓以及血管，实时准确的温度监控是实施安全、有效地肿瘤热疗的必要条件。然而，目前唯一可靠的定量磁共振水质子共振频率测温法,其测温原理导致其准确性受限且无法测量骨小梁等不含水部分的温度；而生物热传导模型对人体结构有所简化，而且在手术中，组织的传热和电介参数会发生变化，此时简单通过术前的模型预测就可能带来较大误差。本项目创新地提出了一种生物热传导模型和磁共振温度成像协同提升的实时温度测定方法，该方法首先利用磁共振结构图像辅助进行正问题的生物热传导建模并计算全区域温度，从而将该模型带入磁共振温度成像代价函数中进行修正，而后再将温度成像得到的含水组织测温结果反馈校正热传导模型参数；以此类推，可以将上述过程迭代进行，两者相互校正，协同作用，提升各自准确性，从而为医生提供准确而实时的椎体骨肿瘤微波消融术中温度监控。

热疗是治疗肿瘤的一种新的有效手段，而对于椎体骨肿瘤这一特殊应用，由于其附近有易受损伤的脊髓以及血管，实时准确的温度监控是实施安全、有效地肿瘤热疗的必要条件。目前流行的定量磁共振水质子共振频率测温法,其测温原理导致其准确性受限且无法测量骨小梁等不含水部分的温度；传统的生物传热模型基于特定的环境以及固定的传热学参数，而在微波消融过程中由于生物组织脱水、变质等等现象，其传热学参数是随着时间而发生变化的。本项目创新地提出了一种生物热传导模型和磁共振温度成像协同提升的实时温度测定方法，该方法首先利用磁共振结构图像辅助进行正问题的生物热传导建模并计算全区域温度，从而将该模型带入磁共振温度成像代价函数中进行修正，而后再将温度成像得到的含水组织测温结果反馈校正热传导模型参数，将上述过程进行迭代，从而得到更高的成像精度以及更好的时间分辨率，在仿体实验上将误差减小到传统磁共振测温方法的1/5左右；同时利用传统的k-空间降采方法以及基于k-空间和图像双领域对抗生成网络的加速重建算法进行磁共振原始数据重建加速，使数据采集与重建能够达到实时更新的要求，从而为医生提供准确而实时的椎体骨肿瘤微波消融术中温度监控。