Automatic image quality evaluation in digital radiography using for-processing and for-presentation images

Area of Science:

• Medical Imaging
• Radiography
• Image Quality Assessment

Background:

• Digital radiography relies on post-processing algorithms to optimize image presentation.
• Understanding the impact of these algorithms on objective image quality metrics is crucial for consistent diagnostic performance.
• Variations in exposure parameters and processing can influence the perceived and measured image quality.

Purpose of the Study:

• To evaluate how digital image post-processing algorithms affect key image quality (IQ) metrics in radiographic images.
• To assess these effects under varying exposure conditions and with different acquisition protocols.
• To compare the IQ metrics between raw (unprocessed) and clinical (processed) radiographic images.

Main Methods:

• Utilized a custom-made phantom and IAEA guidelines for image acquisition.
• Acquired radiographic images using a digital radiography unit, generating both raw and clinical image pairs.
• Analyzed various image quality metrics (IQ-scores) in response to changes in incident air kerma (IAK), tube potential (kVp), filtration, and examination protocols.

Main Results:

• Image quality scores (IQ-scores) demonstrated consistency for repeated exposures across both raw and clinical images.
• Increasing incident air kerma (IAK) and decreasing tube potential (kVp) positively impacted signal difference-to-noise-ratio (SDNR) and detectability index (d') for both image types.
• No significant effect of additional filtration on any IQ metrics was observed. Detectability index (d') was higher in raw images, while other metrics were generally higher in clinical images.

Conclusions:

• Image quality scores differ significantly between raw and clinical images.
• Consistent use of the same image type (raw or clinical) is essential for accurate image quality monitoring and inter-system comparisons.
• Standardization in image type selection is critical for maintaining diagnostic consistency and reliable quality assurance in digital radiography.