Evaluating the relationship between objective and subjective image quality in computed tomography using detectability...

Area of Science:

• Radiology and Medical Imaging
• Biophysics
• Medical Physics

Background:

• Optimal computed tomography (CT) image quality is crucial for accurate medical diagnosis.
• Subjective image quality assessments by radiologists are vital but can be time-consuming.
• Objective metrics are needed to correlate with and predict subjective perceptions of image quality.

Purpose of the Study:

• To evaluate the detectability index (d") as a quantitative metric for computed tomography (CT) image quality.
• To assess the relationship between objective image quality parameters and subjective radiologist assessments.
• To identify key factors influencing perceived CT image quality for protocol optimization.

Main Methods:

• Utilized a 3D-printed anthropomorphic abdominal phantom for CT image acquisition.
• Varied technical parameters: tube voltage, current, slice thickness, and reconstruction kernel.
• Measured objective metrics (SNR, CNR, edge sharpness) and calculated the detectability index (d").
• Correlated objective metrics with subjective assessments from experienced radiologists.

Main Results:

• A strong positive correlation was found between the detectability index (d") and subjective image quality ratings.
• Multiple linear regression identified d", object size, and noise power spectrum peak spatial frequency as significant predictors of perceived quality.
• Objective measures, particularly d", effectively predicted radiologist-perceived image quality.

Conclusions:

• The detectability index (d") is a reliable and quantitative predictor of subjective CT image quality.
• Integrating d" into routine CT evaluations can help align technical parameters with clinical expectations.
• This approach supports improved CT protocol optimization for enhanced diagnostic performance.