Published-IP

This disclosure covers the 3D volumetric algorithm for detecting and registering a stereotactic localizer in a series of CT and MRI images. This method is different from other techniques by the working across the complete series of images, while the traditional method only uses a single image slice. This provides a more robust solution by reducing the errors caused by noise and missing data. The 3D volumetric detection allows for enhanced troubleshooting and error correction that cannot be obtained using that single slice method. Such as bent rod detection, patient motion, missing rods, air bubbles, etc.

The basic algorithm flow is outlined as follows:
1) Detection of potential rod locations
2) Construct lines from potential rod locations
3) Detect localizer type and initial transformation
4) Perform detailed line detection using localizer model and transform
5) Refine transform using ICP technique
6) Calculate error measures

This algorithm has been reviewed for patient infringements as part of the Stealth Foundation project.