A fast CT and CT-fluoroscopy registration algorithm with respiratory motion compensation for image-guided lung...

Area of Science:

• Medical Imaging
• Interventional Radiology
• Image-guided therapy

Background:

• CT-fluoroscopy (CTF) guides lung interventions but involves frequent scans, increasing radiation and limiting 3-D information.
• Current CTF methods provide limited 2-D data, posing challenges for volumetric anatomical visualization and precise needle guidance.

Purpose of the Study:

• To develop a fast CT-CTF deformable registration algorithm for 3-D image guidance during percutaneous lung interventions.
• To enable accurate warping of preprocedural CT scans onto intraprocedural CTF images, reducing intraoperative scanning and enhancing anatomical visualization.

Main Methods:

• A novel fast CT-CTF deformable registration algorithm utilizing 2-D B-Spline for transverse plane deformation and smoothness constraints for z-direction.
• Incorporation of a respiratory motion compensation framework for enhanced registration accuracy.
• Parallel implementation for registration completion within seconds, coupled with electromagnetic tracking for needle position superimposition.

Main Results:

• The algorithm successfully warped inhale preprocedural CT scans onto intraprocedural CTF images, providing 3-D anatomical information.
• Experiments with simulated and real lung cancer biopsy data demonstrated satisfactory registration accuracy.
• The method facilitates 3-D image guidance during breath-holding, improving needle placement accuracy.

Conclusions:

• The proposed fast CT-CTF deformable registration algorithm offers an effective solution for 3-D image guidance in lung interventions.
• This approach has the potential to reduce procedure time, complication rates, and radiation exposure for both patients and clinicians.
• Accurate 3-D anatomical information derived from CTF guidance can significantly improve the precision and safety of percutaneous lung procedures.