Multiparametric Characterisation of a Recursive Filtration Algorithm on a Commercial Angiographic Equipment.

Area of Science:

• Medical Imaging
• Digital Fluoroscopy
• Image Processing

Background:

• Digital fluoroscopy provides high temporal resolution but suffers from significant noise.
• Recursive filtration enhances signal-to-noise ratio (SNR) by integrating information across frames.
• This technique may compromise image quality through motion blur and contrast reduction.

Purpose of the Study:

• To evaluate the performance of recursive filtering in a commercial angiographic system.
• To assess the impact of undisclosed filter parameters, such as the K-factor.
• To compare manual K-factor settings with automatic configurations.

Main Methods:

• Acquisition of fluoroscopic images using a rotating phantom with varying contrast objects.
• Image quality assessment via noise power spectrum (NPS), contrast-to-noise ratio (CNR), and lag analysis.
• Development of a model to estimate the effective K-factor based on NPS characteristics.

Main Results:

• Higher K-factors significantly reduced noise but increased lag and decreased contrast.
• Automatic filtering modes achieved noise suppression comparable to high K-factors while preserving contrast and minimizing motion artifacts.
• Estimated K-factors in automatic modes demonstrated adaptive behavior, varying with temporal frequency.

Conclusions:

• Understanding undisclosed filter settings is crucial for optimizing digital fluoroscopy protocols.
• Adaptive filtering strategies can enhance imaging performance and patient safety in clinical settings.
• Automatic recursive filtering offers a balanced approach to noise reduction and image quality preservation.