Curvelet based contrast enhancement in fluoroscopic sequences.

Area of Science:

• Medical Imaging
• Image Processing
• Radiology

Background:

• Image-guided interventions require frequent X-ray imaging, necessitating dose reduction strategies.
• Low-dose X-ray sequences suffer from high noise and low contrast, hindering visualization.
• Efficient image processing methods are crucial for optimizing diagnostic quality in low-dose imaging.

Purpose of the Study:

• To develop an advanced denoising and feature enhancement method for low-dose X-ray image sequences.
• To enable significant radiation dose reduction in image-guided interventions without compromising image quality.
• To improve the visualization of low-contrasted structures in medical imaging.

Main Methods:

• A three-step denoising and enhancement approach using the curvelet transform.
• Application of recursive temporal filtering followed by spatial filtering of curvelet coefficients.
• Inclusion of a line enhancement technique in the curvelet domain for feature detection.

Main Results:

• The proposed method effectively denoises sequences while preserving low-contrasted structures.
• A significant 50% reduction in radiation dose was demonstrated on synthetic and real low-dose sequences.
• The technique enhances the visibility of thin structures crucial for diagnosis.

Conclusions:

• The curvelet transform-based method offers a viable solution for dose reduction in image-guided interventions.
• This approach maintains and enhances image quality, crucial for accurate medical diagnosis.
• The developed technique facilitates safer and more effective medical imaging procedures.