C11522: Automatic 3D-2D Registration Methods for Image-guided SurgeryNovelty:
The current invention is a method for target visualization under 3D-2D imaging.
Value Proposition:
Currently the use of fluoroscopic level counting is associated with errors which can give an unreliable estimate to surrounding structures. The invention reduces the time of surgery and has the potential to update the registration process in a near real-time manner. The demonstrated accuracy and simplicity of the method combines with the following advantages:
• Image guidance for structures in the vicinity of the target vertebrae
• Reduced surgical time
• Faster (automatic) localization of target structures within fluoroscopic images
• Reduced reliance on fluoroscopy and, therefore, lower levels of ionizing radiation to the patient and intraoperative staff
• Increased accuracy (reduced rate of errors) in localizing target structures
Technical Details:
Johns Hopkins researchers have developed a method for automatically computing locations of anatomical and other structures within a 2D projection image from a 3D CT scan. The structures could be any 3D structure of interest tumors, anatomical landmarks, vessels, nerves, etc. A 3D-2D registration algorithm provides information of the target structure location in the intraoperative x-ray image acquired with a mobile C-arm. The system includes a graphical processing unit which generates a digitally reconstructed radiograph (DRR) from the known surgical protocol and system geometry. A similarity measurement compares information from the DRR (floating image) and the preprocessed x-ray projection images (fixed image) to visualize the edges and boundaries of the structure under examination such as a vertebrae. The result is an image overlaid onto another which helps determine the location of the 3D object.
Looking for Partners:
To develop and commercialize the technology as software method of 3D-2D registration for automatic structure localization.
Stage of Development:
Prototype
Data Availability:
Under CDA/NDA
Publications/Associated Cases:
1. Phys Med Biol. 2012 Sep 7;57(17):5485-508;
2. SPIE Medical Imaging 2012, Volume 8316;
3. Automatic Localization of Vertebral Levels in C-Arm Fluoroscopy: Pre-Clinical Evaluation of the LevelCheck Algorithm;
4. Automatic Localization of Target Vertebrae in Spine Surgery Using Fast CT-to-Fluoroscopy (3D-2D) Image Registration;
5. The 2013 Spine Technology Award;
6. PMB Article: 3D-2D Registration Enables Surgical Guidance