Needle Placement under X-ray Fluoroscopy using Perspective Invariants

This paper presents an approach for image-based guidance of a surgical tool towards multiple targets from fixed or variable entry points. The method is based on visual servoing. It requires no prior calibration or registration. By taking advantage of projective invariants, precise needle alignment to a target can be achieved in a fixed number (12) of iterations. Alignment to n targets can be performed in 6*(n+1) iterations. Elements of error analysis and a discussion of the “optimal” placement of the planes used in the method are given. We also show how the approach can be used to estimate the entry point and orientation to reach an anatomic target while passing through a given anatomic landmark.