Instrument visual tracking method for laparoscopic minimally invasive surgery

Abstract

The invention discloses a surgical instrument visual tracking method for laparoscopic minimally invasive surgery, which is based on a deep learning method and comprises the following steps: firstly, a surgical instrument area is detected and segmented from a surgical video stream and extracting local features on the premise of not performing preoperative marking on a surgical instrument; the detection and positioning efficiency and the feature extraction accuracy of the surgical instrument are greatly improved; meanwhile, image enhancement processing is carried out on the region of interest by utilizing an image filtering algorithm and region screening, so that the recognition precision of a tracking target point is improved; secondly, a 2D-3D conversion relation between the surgical instrument and the target area is calculated through modeling and mathematical expression so as to determine depth perception information of an end effector of the surgical instrument and the target area; and finally, a real-time tracking program of the target point is realized by using a mask algorithm. A data set is established through clinical medicine guidance and used for training and testing a surgical instrument detection model, automatic detection and tracking of the laparoscopic surgical instrument are achieved, and the model is more practical.

Description

Technical field [0001] The present invention relates to the field of laparoscopic microsurgical surgery, and in particular, there is a device visual tracking method for laparoscopic microsuroscope. Background technique [0002] Minimally invasive surgery (MIS) is a new surgical technology. Doctors do not have to operate directly to patient lesions during surgery, by inserting the patient's abdominal cavity, while projection of laparoscopic video streams to the display The procedure process. Relative to traditional open surgery, laparoscopic micro-surgical wound is fast, postoperative nursing is simple, is currently the development trend of chest and abdominal intraoperative surgery, but this indirect operation will make medical devices and depth perceived information in surgical instruments and lesions. These surgery remain very challenging for surgeons. [0003] Laparoscopic microsuroscope is to determine and obtain relative positions of the surgical instrument and the target ob...

AI Technical Summary

Problems solved by technology

[0007] The object of the present invention is to provide a kind of instrument visual tracking method for laparoscopic minimally invasive surgery, in order to at least solve the depth, direction and relative pose perception error of surgical instrument and target area in existing laparoscopic minimally invasive surgery. Large, and one or more problems such as low detection efficiency of surgical instruments, insufficient positioning accuracy, and unreliable real-time tracking delay

Images