An interactive task-based method for the avoidance of metal artifacts in CBCT

Area of Science:

• Medical Imaging
• Orthopedic Surgery
• Radiology

Background:

• Intraoperative cone-beam CT (CBCT) is crucial for 3D validation in orthopedic and trauma surgeries.
• Metal artifacts severely degrade CBCT image quality, limiting clinical utility.
• Existing metal artifact avoidance (MAA) methods lack clinical integration and workflow compatibility.

Purpose of the Study:

• To develop an interactive method for predicting and visualizing metal artifacts in CBCT.
• To enable clinicians to select optimal scanning trajectories based on artifact distribution.
• To improve the clinical applicability of MAA techniques.

Main Methods:

• A novel method to compute spatial artifact distribution and calibrated strengths for tilted trajectories.
• Interactive visualization of predicted artifacts overlaid on C-Arm tilt.
• Evaluation in a human cadaver study comparing interactive vs. global metric optimization.

Main Results:

• Both global metrics and the proposed visualization significantly improved image quality over standard scans.
• The interactive visualization provided superior subjective image quality compared to global metrics.
• The method allows task-specific optimization to avoid artifacts in regions of interest.

Conclusions:

• A practical and effective interactive MAA method was developed to reduce metal artifacts.
• The approach provides spatially resolved artifact prediction and calibrated strength grading.
• This represents a significant step towards clinical implementation of MAA for improved surgical validation.