A novel method for artery detection in laparoscopic surgery

Abstract

One of the most important difficulties in laparoscopic surgery is intraoperative blood vessel detection. An accidental injury to a blood vessel may cause serious complications and could result in changing from a laparoscopic procedure to open surgery. Moreover, differentiating arteries from veins is necessary in all surgical cases. In this study we evaluate a new image-processing method for artery detection that would be useful during laparoscopic and endoscopic procedures. It is possible to install the program on any ordinary laparoscopy set and it displays the artery's region on the monitor. This method uses the artery's pulse to detect an artery and distinguish it from veins. By subtracting the systolic and diastolic images, the change regions are detected and shown on a monitor. The performance of this method in detecting arteries in simulation and in real laparoscopic surgery is evaluated. Artery detection in different pulse rates, different artery depths, and different blood pressures is tested via the simulation phase. It is also tested in two laparoscopic surgeries, one on a kidney and one on a stomach. In simulation phase the method can correctly detect all arteries that are not too deep and can move superficial tissues with zero false-negative and false-positive rates. In real laparoscopy, the false-positive rate was 8% and the false-negative rate was 5%. This method is a noninvasive, reliable, and cost-effective technique to detect artery regions, even if some of them are covered with fat or other tissues, while suppressing veins and other tissues.