An Improved Visual Auxiliary Algorithm for the Vascular Interventional Surgical Robot based on Neural Network

Given the various shortcomings of traditional interventional surgery, it has become a trend to develop vascular interventional robots to assist doctors in completing the surgery. However, the existing visual auxiliary for vascular interventional robots rely on doctors’ experience to make judgments based on images. Therefore, our team proposed a method of transforming image coordinates and spatial coordinates based on the visual auxiliary of existing platforms. This paper shows the platform briefly at first. Then we introduce the traditional image coordinate transformation and use the neural network to simplify calculation and parameter preparation process based on the traditional way. Finally, the accuracy and practicability of the transformation between the vision-assisted image coordinates and the space coordinates of the vascular interventional surgical robot based on the neural network are proved through experiments. We can apply this algorithm to the visual assistant and give detailed numerical feedback to the doctor intuitively, greatly improving the accuracy of the operation.