Enhancing Single-Plane Fluoroscopy: A Self-Calibrating Bundle Adjustment for Distortion Modeling

Area of Science:

• Medical Imaging
• Geometric Calibration
• Image Processing

Background:

• Single-plane fluoroscopy systems are widely used clinically.
• These systems are prone to geometric distortions from components, assembly, and magnetic fields.
• Accurate geometric calibration is crucial for reliable quantitative analysis and diagnostics.

Purpose of the Study:

• To investigate a self-calibrating bundle adjustment for correcting geometric distortions in fluoroscopic imaging.
• To assess the impact of calibrated imagery on diaphragmatic motion analysis and hemidiaphragm paralysis diagnosis.
• To explore the potential of calibrated imagery in surgical navigation.

Main Methods:

• A controlled experiment using three Philips Easy Diagnost R/F Systems.
• Collection of a highly redundant image network (18-22 images) of a target field with 121 pre-surveyed tantalum beads.
• Application of bundle adjustment to estimate distortion coefficients and calibrate imagery.

Main Results:

• The modeling process estimated five to eight distortion coefficients per system.
• An 83-85% improvement in image point precision (model fit) was achieved.
• An 85-95% improvement in 3D object reconstruction accuracy was observed after calibration.

Conclusions:

• The self-calibrating bundle adjustment effectively corrects geometric distortions in single-plane fluoroscopy.
• This calibration enhances the accuracy and reliability of fluoroscopic imaging.
• Improved imaging quality has significant potential for medical diagnostics, treatment, and surgical navigation.