Markerless surgical robotic system for intracerebral hemorrhage surgery.

Conventional intracerebral hemorrhage (ICH) surgery uses a stereotactic frame to access an intracerebral hematoma. Using a stereotactic frame for ICH surgery requires a long preparation time. In order to resolve this problem, we propose a markerless surgical robotic system. This system uses weighted iterative closest point technology for surface registration, hand-eye calibration for needle insertion, and 3D surface scanning for registration. We need calibration to integrate the technologies: calibration of robot and needle coordinates and calibration of 3D surface scanning and needle coordinates. These calibrations are essential elements of the markerless surgical robotic system. This system has the advantages of being non-invasive, a short total operation time, and low radiation exposure compared to conventional ICH surgery.

Robust surface reconstruction of teeth from raw pointsets.

This paper discusses a 3D robust dental surface reconstruction method. The surface reconstruction method usually consists of the following stages: mesh generation, surface registration, and surface merging. A large amount of noise will be accumulated after these three processes especially for the dental model with many sharp features of high curvature. To obtain an accurate dental surface, this paper first strengthens the mesh generation and the surface registration using optimal parameters. Then, a 3D robust dental surface reconstruction is performed by iteratively executing the smallest univalue segment assimilating nucleus filtering method and the surface attraction method. Finally, virtual scanning and real scanning models are introduced to validate the accuracy of the presented method. The experiments show that our method can robustly reconstruct the 3D dental model.