Heterogeneous Stitching of X-ray Images According to Homographic Evaluation

Area of Science:

• Medical Imaging
• Orthopedic Surgery
• Computer Vision

Background:

• C-arm X-ray systems are crucial for intraoperative orthopedic surgery, aiding in fracture reduction and alignment verification.
• The narrow field-of-view of C-arm systems necessitates stitching multiple images to create a comprehensive panoramic view.
• Accurate feature matching in overlapping regions is critical for successful image stitching but is often challenging due to varying overlap properties.

Purpose of the Study:

• To develop and validate a robust method for heterogeneous stitching of C-arm X-ray images.
• To enhance the accuracy and density of feature matching in limited overlap regions for improved panoramic image construction.
• To overcome limitations of conventional stitching methods in orthopedic surgical applications.

Main Methods:

• Proposed a heterogeneous stitching approach based on homographic evaluation of overlap regions.
• Utilized integrated feature detection to estimate and evaluate homography for accurate feature matching.
• Implemented a re-estimation process for homography using local image regions when initial estimations were inaccurate.

Main Results:

• Successfully achieved stitching of C-arm X-ray images by estimating optimal homography for each image.
• Demonstrated robust panoramic X-ray image construction through phantom and ex-vivo experiments.
• The proposed method showed improved performance compared to conventional stitching techniques.

Conclusions:

• The developed heterogeneous stitching method provides a robust solution for creating panoramic X-ray images from C-arm data.
• This technique enhances intraoperative visualization in orthopedic surgery, facilitating more accurate fracture reduction.
• The findings suggest a significant advancement in medical imaging for surgical guidance and verification.