Objective: Compared with traditional surgery, laparoscopic surgery offers the advantages of smaller scars and rapid recovery and has gradually become popular. However, laparoscopic surgery has the limitation of low visibility and a lack of touch sense. As such, a physician may unexpectedly damage blood vessels, causing massive bleeding. In clinical settings, Doppler ultrasound is commonly used to detect vascular locations, but this approach is affected by the measuring angle and bone shadow and has poor ability to distinguish arteries from veins. To tackle these problems, a smart blood vessel detection system for laparoscopic surgery is proposed.
Methods: Based on the principle of near-infrared spectroscopy, the proposed instrument can access hemoglobin (HbT) parameters at several depths simultaneously and recognize human tissue type by using a neural network.
Results: Using the differences in HbT and StO2 between different tissues, vascular and avascular locations can be recognized. Moreover, a mechanically rotatable stick enables the physician to easily operate in body cavities. Phantom and animal experiments were performed to validate the system's performance.
Conclusion: The proposed system has high ability to distinguish vascular from avascular locations at various depths.
Keywords: Laparoscopic surgery; hemoglobin parameters; near infrared spectroscopy; neural network; vascular location.