Development of real-time color intravascular ultrasound using Wavelet Analysis

背景和目的：小波分析是一种新的数学模型，用于评估时间序列信号几何轮廓的局部变化。我们先前报道，血管内超声(IVUS)视频图像的小波分析可能有助于检测体内有动脉粥样硬化斑块的富脂区域。本研究的目的是开发基于小波分析的实时彩色血管内超声检测系统。方法：用40 MHz的IVUS导管采集32块福尔马林固定的非钙化动脉粥样硬化斑块的IVUS信号。用Daubechies-2小波对图像进行小波分析，得到小波尺度x-y平面上重构的小波与IVUS导管距离的相关系数。结果：组织学检查显示26个斑块中有脂质斑块。在进行小波分析时，鸟瞰显示了与其他类型的斑块相比，载脂斑块中明显不同的模式。利用该小波分析可以检测出载脂斑块，其敏感性为81%(26/32)，特异性为80%。

Background and Objects: Wavelet analysis is a new mathematical model for assessing local changes in geometrical profile of time-series signals. We previously reported that Wavelet analysis of video-images of intravascular ultrasound (IVUS) might be useful for detecting lipid-rich area with atherosclerotic plaque in vivo. The purpose of this study is the development of real-time color intravascular ultrasound using Wavelet Analysis. Methods: IVUS signals were acquired from 32 formalin-fixed noncalcified atherosclerotic plaques from human necropsy by a 40 MHz IVUS catheter. Wavelet analysis of these images was performed with Daubechies-2 wavelet to obtain the correlation coefficient with the wavelet reconstructed over the x-y plane of the wavelet scale and the distance from IVUS catheter. Results: Histological examination revealed lipid-laden segment in 26 plaques. When performing a wavelet analysis, the bird eye view revealed an apparently different pattern in lipid-laden plaque compared to other types of plaque. Using this wavelet analysis, the lipid-laden plaque could be detected with sensitivity of 81%( 26/32) and specificity of 80%.