基于近红外血管成像的浅静脉三维定位研究

Three-dimensional localization of superficial veins based on near-infrared angiography

目的:基于近红外血管成像进行浅静脉三维定位,以为疫情时代下的医护人员进行静脉穿刺提供参考。方法:首先设计并搭建近红外静脉图像采集系统;其次对近红外静脉图像进行对比度受限自适应直方图均衡化(contrast limited adaptive histgram equalization,CLAHE)处理、二值化处理、形态学处理和细化处理,确定血管的二维位置和血管走向;最后通过建立多波长近红外静脉深度恢复数学模型计算血管深度,并与超声显示血管深度进行对比。结果:在871个样本点中,该方法计算的血管深度与超声显示的血管深度误差≤0.021 cm的达到97.82%。结论:提出的方法能够恢复二维场景中丢失的深度信息,从而实现对浅静脉的三维定位。

Objective To realize 3D localization of superficial veins based on near-infrared(NIR) angiography to facilitate venipuncture in the epidemic era. Methods Firstly, a near-infrared vein image acquisition system was designed and de-veloped, and then the two-dimensional location and orientation of the vessels were determined with contrast limited adaptive histogram equalization(CLAHE), binarization processing, morphology treatment and refine processing of NIR vein images,finally the vessel depths were calculated by building a multi-wavelength NIR vein depth recovery mathematical model and were compared with those shown by ultrasound. Results The error not higher than 0.021 cm between the vessel depths calculated by this method and shown by ultrasound reached 97.82% in 871 sample sites. Conclusion The proposed method is able to recover the lost depth information in 2D scenes, thus enabling 3D localization of superficial veins.