Development of blood vessel search system using near-infrared light for laparoscopic surgery

Our objective of this study is to develop a miniature and high accuracy automatic 3D blood vessel searching system, which will be introduced in the laparoscopic operation with the minimally invasive surgery. Now, the conventional optical system used in the laparoscopic surgery has many difficulties of blood vessel imaging and detection, because the peripheral bio-tissue located around the blood vessel disturbs the light wave propagation, disperses and refracts. Consequently, only unclear image is obtained. We develop a new blood vessel detecting system by using Near-Infrared (NIR) light, two CMOS camera modules and a comprehensive image processing technique, which is implemented in the laparoscope pipe with 25mm in diameter. We adopt the stereo method for the searching system to determine 3D blood vessel location. The blood vessel visualization system adopts hemoglobin’s absorption characteristics of the NIR light, which has high permeability for the bio-tissue and absorbency for the hemoglobin. A sharpening process is employed to improve the image quality of original ones, through the LoG filter and the un-sharp-mask processing. 2D location of the blood vessel is calculated from luminance distribution of the image and its depth is calculated by the stereo method. A validity of our blood vessel visualization and 3D detecting system was examined through the comparison with the imaging and detecting the results of organization phantoms, which embedded at known depths under the surface. Experimental results of depth obtained by our detecting system showed good agreements with the given depths, and the availability of this system is confirmed.