Optoacoustic (OA) angiography allows high-contrast three-dimensional (3D) visualization of hemoglobin-containing structures ranging from micrometers to millimeters. However, due to the large amount of 3D data acquired by modern high-throughput OA systems the resulting OA vasculature images might be difficult to analyze visually. This problem is especially relevant for monitoring of angiogenesis of experimental tumors, which blood vessels tend to be smaller and more tortuous compared to vasculature of healthy tissue. In this paper a novel algorithm for OA image processing is proposed to quantify vessel structure parameters automatically. The algorithm is based on creation of vasculature graphs which parameters (lengths of branches, number of branches, etc) can serve as a numerical characterization of vasculature: vessel density, vessel length, etc. The results of testing the developed algorithm on numerical simulation phantoms and in vivo OA images of tumor models in a mouse demonstrate a statistically significant difference of all the extracted parameters for tumor and normal tissue. The results show a high potential of the proposed approach for OA angiography in different applications including clinical and experimental oncology.

光声（OA）血管造影可以对从微米到毫米的含血红蛋白的结构进行高对比度的三维（3D）可视化。但是，由于现代高通量OA系统获取的大量3D数据，因此可能难以在视觉上分析所得的OA脉管系统图像。这个问题与监视实验性肿瘤的血管生成特别相关，与健康组织的脉管系统相比，该血管的血管倾向于更小且更弯曲。本文提出了一种新的OA图像处理算法，可以自动量化血管结构参数。该算法基于创建的脉管系统图，该参数（分支的长度，分支数等）可以用作脉管系统的数字表征：血管密度，血管长度等。小鼠体内肿瘤模型的体内OA图像显示了所有提取的肿瘤和正常组织参数的统计上的显着差异。结果表明，所提出的OA血管造影方法在包括临床和实验肿瘤学在内的不同应用中具有很大的潜力。