基于同步辐射CT成像的乳腺癌微血管的定量研究

发布时间：2018-12-16 05:07

【摘要】：肿瘤新生血管对恶性肿瘤的发生、发展和转移具有重要作用,肿瘤微血管网络的形态结构也反映着肿瘤的生长状态。而目前医疗领域所使用的成像方法,如CT、MRI、超声,仅能分辨100微米以上的血管,无法观察到肿瘤新生血管,给肿瘤的早期诊断带来了很大的困难。同步辐射光源作为新兴的成像手段,有着高强度、高耀度等优势,分辨率较传统成像方法有了近千倍的提高。microfil作为研究微循环的常用造影剂,能够分辨直径20微米以下的血管(Michael Marxen et al., 2003),二者的结合解决了传统影像技术分辨率不足的问题,为肿瘤新生血管结构的精确定量提供可能。本研究基于4T1乳腺癌的BALB/c雌鼠模型,利用microfil作为肿瘤微血管造影剂,在上海光源生物医学成像线站13W进行离体肿瘤样本的CT成像,获取不同时间点的肿瘤微血管数据(2GB~5GB),使用常规的滤波反投影算法进行重建后体绘制,观测肿瘤微血管形态的变化趋势。另外,本文提出了针对本课题大数据量的肿瘤血管的分块加载的双阈值分割方法。为了研究血管生长与生长区域的关系,将肿瘤按离中心点的距离进行分区,由外到内依次为外围区域、过渡区域和中心区域。结果表明,利用同步辐射CT造影成像能够观察到直径为20微米左右的肿瘤新生血管;由血管密度和分形维度的统计结果来看,恶性肿瘤生长演变迅速,在急剧增殖不久后便迅速坏死、转移,且这一转折点出现在第25天。25天以前肿瘤微血管较少,肿瘤细胞增殖速度较小;25天肿瘤微血管密度达到最大,由分形维度反映出的复杂度也最大;25天以后肿瘤体积增速明显升高,达到最大增速,但微血管密度明显下降,且伴有肺部和淋巴系统的转移。实验结果反映出新生血管对于肿瘤生长、转移的预示作用。虽然微血管密度的变化趋势与分形维度吻合,然而由于肿瘤微血管分布的不均匀性,导致分区后的微血管密度的统计存在较大方差,使得其可信度大大降低。由此也反映出微血管密度的测定与采样区域密切相关,因此其客观性和稳定性不及分形维度。研究结果表明利用同步辐射光源观察肿瘤微血管的实验方法和自主设计的分块加载的双阈值分割方法具有可行性,分形维度对于肿瘤微血管定量研究具有重要作用。在今后的研究中,我们将进一步优化分区,拓展分形维度在肿瘤微血管定量中的运用。
[Abstract]:Tumor neovascularization plays an important role in the occurrence, development and metastasis of malignant tumors, and the morphological structure of tumor microvascular network also reflects the growth state of tumor. At present, the imaging methods used in the medical field, such as CT,MRI, ultrasound, can only distinguish more than 100 microns of blood vessels, and can not observe tumor neovascularization, which brings great difficulties to the early diagnosis of tumor. As a new imaging method, synchrotron radiation source has many advantages such as high intensity, high flare and so on. The resolution of synchrotron radiation light source is nearly a thousand times higher than that of traditional imaging method. Microfil is used as a contrast medium to study microcirculation. The combination of (Michael Marxen et al., 2003 with diameter less than 20 渭 m solves the problem of insufficient resolution of traditional image technology and provides the possibility for accurate quantification of tumor neovascularization. In this study, based on the BALB/c female model of 4T1 breast cancer, microfil was used as tumor microvessel contrast medium to perform CT imaging of isolated tumor samples at 13 W of Shanghai Light Source Biomedical Imaging Line Station. The tumor microvessel data (2GB~5GB) at different time points were obtained, and then reconstructed by conventional filtering back-projection algorithm, and the trend of tumor microvessel morphology was observed. In addition, this paper proposes a double threshold segmentation method for tumor vessels with large amount of data. In order to study the relationship between angiogenesis and growth region, the tumor was divided according to the distance from the center to the center, and the outer region, the transitional region and the central region were sequentially divided from outer to inner. The results showed that tumor neovascularization with diameter of 20 渭 m could be observed by synchrotron radiation (CT) imaging. According to the statistical results of vascular density and fractal dimension, malignant tumors grow rapidly, necrosis and metastasis occur soon after rapid proliferation, and this turning point appears on the 25th day. The proliferation rate of tumor cells was small; The tumor microvessel density reached the maximum at 25 days, and the complexity reflected by fractal dimension was the largest. After 25 days, the tumor volume growth rate increased significantly, reached the maximum growth rate, but the microvessel density decreased significantly, and accompanied by lung and lymphatic system metastasis. The results reflect the predictive effect of neovascularization on tumor growth and metastasis. Although the variation trend of microvessel density is consistent with fractal dimension, the statistical variance of microvessel density after partition exists because of the non-uniformity of tumor microvessel distribution, which greatly reduces the reliability of microvessel density. It also shows that the measurement of microvessel density is closely related to the sampling area, so its objectivity and stability are not as good as fractal dimension. The results show that the experimental method of using synchrotron radiation light source to observe tumor microvessels and the self-designed dual-threshold segmentation method with block loading are feasible. Fractal dimension plays an important role in quantitative study of tumor microvessels. In the future research, we will further optimize the partition and expand the application of fractal dimension in tumor microvessel quantification.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R737.9;R318

【共引文献】