Research On Theory And Key Technology Of Radiation-free Navigation For Minimally Invasive Spinal Surgery

The incidence of spinal cord diseases trends to raise year by year.Surgical treatments are needed in serious conditions.Minimally invasive spinal surgeries are becoming one of the most mainstream methods which relies heavily on effective surgical navigation and positioning.Intraoperative X-ray fluoroscopy is always regarded as the main way of surgical navigation and positioning,with the advantages of high-performance imaging and mobility.Instead of once fluoroscopy,repeated has become necessary owing to the demand of better precision.Thus,long-term radiation exposure and security has aroused more and more concern.Reduce and remove the risk of radiation exposure while improve the performance of surgical navigation system,have become a driving force for spreading minimally invasive spinal surgery.They are deemed as the technical bottleneck of high performance and machine assisted surgical treatments which needs for breakthrough in priority.Given the above background,supported by the Provincial Key R&D Program of Zhejiang Province,this article proposed the theory and key technology research of radiation-free navigation for minimally invasive spinal surgery.On the basis of systematic understanding of relevant research status and development trend,the main technical route based on ultrasonic imaging was established for radiation-free navigation during minimally invasive spinal surgeries.This research was carried out from three aspects,which are machine assisted self-adaptive ultrasound scanning imaging,ultrasonic image processing based on machine learning,ultrasonic coronal reconstruction and image matching for navigation and positioning.The prototype was integrated and target segments was positioned by the way of radiation-free navigation during minimally invasive spinal surgeries.In the meantime,the spinal phantoms are designed and used to verify the feasibility and effectiveness of this work.The main content is embodied as follows:(1)Project of radiation-free navigation system and design of human spine phantom.Aimed at the current issues in navigation of spinal surgeries represented by radiation exposure,medico engineering cooperation was established.Researchers designed the technical route of radiation-free navigation based on machine assisted self-adaptive ultrasound scanning imaging.Self-adaptive Attitude-adjusting Mechanism and human spine phantom were designed and developed independently.(2)Machine assisted self-adaptive ultrasound scanning imaging for human spine.Analyzed the key points when sonographers deal with human spine imaging in the situation of prone position under general anesthesia.Developed the machine assisted self-adaptive ultrasound scanning imaging,focus on machine vision planning of ultrasonic scanning path on human spine and optimization of ultrasonic scanning force and speed.With eliminating the factors of human spine ultrasound scanning and imaging,compared with sonographers,higher performance ultrasonic images were able to obtained.(3)Ultrasonic image processing based on machine learning for human spine.After the preprocessing of noise elimination and contour extraction on ultrasonic images,the Unet model was established for structural feature extraction of human spine on the images.By training and testing,the machine learning analysis mechanism was promoted to extracting structural characteristic information from ultrasonic images of human spine.(4)Ultrasonic coronal reconstruction and image matching for navigation and positioning of human spine.Accomplish the ultrasonic coronal 3D reconstruction with high performance by optimizing the imaging parameters.Given spatial position information to spinal ultrasonic images by the calibrated magnetic positioning system.Process the ultrasonic coronal reconstruction images and the preoperative CT images into 3D point cloud.Match the 3D point cloud for navigation and positioning.(5)Radiation-free navigation system and clinical test for minimally invasive spinal surgeries.Design the man-machine friendly interactive interface,which pools the function of ultrasonic scanning and imaging,processing,reconstruction,matching and positioning together.Integrate the navigation prototype and conduct the clinical test of target segments positioning on patients with thoracolumbar spinal cord diseases.The advantages of radiation-free and lower operating time were evaluated,which showed a promising prospect in minimally invasive spinal surgery.