Purdue University, Weldon School of Biomedical Engineering, West Lafayette, Indiana 47907, USA.
J Biomed Opt. 2012 Sep;17(9):96010-1. doi: 10.1117/1.JBO.17.9.096010.
Photoacoustic microscopy using vibrational overtone absorption as a contrast mechanism allows bond-selective imaging of deep tissues. Due to the spectral similarity of molecules in the region of overtone vibration, it is difficult to interrogate chemical components using photoacoustic signal at single excitation wavelength. Here we demonstrate that lipids and collagen, two critical markers for many kinds of diseases, can be distinguished by multispectral photoacoustic imaging of the first overtone of C-H bond. A phantom consisting of rat-tail tendon and fat was constructed to demonstrate this technique. Wavelengths between 1650 and 1850 nm were scanned to excite both the first overtone and combination bands of C-H bonds. B-scan multispectral photoacoustic images, in which each pixel contains a spectrum, were analyzed by a multivariate curve resolution-alternating least squares algorithm to recover the spatial distribution of collagen and lipids in the phantom.
基于振动泛频吸收的光声显微镜技术可实现深层组织的键选择性成像。由于泛频振动区域内分子的光谱相似性,因此很难在单一激发波长下使用光声信号来探测化学组分。本文展示了利用 C-H 键的第一泛频的多光谱光声成像技术,可以区分脂质和胶原蛋白,这两种物质是许多疾病的关键标志物。本文构建了一个包含鼠尾肌腱和脂肪的仿体来验证该技术。实验扫描了 1650nm 到 1850nm 之间的波长,以激发 C-H 键的第一泛频和组合频带。通过多元曲线分辨-交替最小二乘法算法对包含每个像素光谱的 B 扫描多光谱光声图像进行分析,以恢复仿体中胶原蛋白和脂质的空间分布。