A Binocular Vision-Guided Puncture Needle Automatic Positioning Method

In the traditional Magnetic Resonance Imaging(MRI)-guided breast puncture surgery, doctors perform puncture surgery manually based on the MRI images and experience. The puncture needle may not be able to reach the location of the lesion accurately due to the inexperience or fatigue of the doctor. In this paper, a binocular vision-guided automatic positioning method for breast puncture needles is proposed. First, the combined image processing algorithm and the three-frame difference method are used to detect the needle tip; then, the puncture needle is guided to the target position by position-based visual servoing. The experimental results show that the average error of feature point detection is within 3 pixels, and the final positioning error is within 2.5mm. Compared with the traditional manual positioning method, the positioning accuracy of this method is higher, and the puncture process is also more stable, which improves the efficiency of surgery.