Sci-Sat AM(1): Imaging-07: Open field normalization: How to avoid inflation to MTF and DQE values caused by...

Area of Science:

• Medical Imaging Physics
• Image Quality Assessment

Background:

• Accurate calculation of imaging system performance metrics like detective quantum efficiency (DQE) is crucial.
• Traditional methods using zero-frequency normalization of the modulation transfer function (MTF) can introduce significant errors.
• These errors arise from using a finite region of interest (ROI) in image data analysis.

Purpose of the Study:

• To introduce and validate a novel open-field normalization technique for MTF calculation.
• To demonstrate how this technique prevents common DQE calculation errors.
• To address the inaccuracies caused by zero-frequency normalization in DQE assessments.

Main Methods:

• Modeled zero-frequency and open-field normalization techniques.
• Derived measured MTF, noise power spectrum (NPS), and DQE using simulated 1D images with Gaussian blur.
• Utilized varying ROI sizes (1-10 cm) for comparative analysis.

Main Results:

• Finite ROIs cause MTF truncation and spectral leakage in MTF and NPS.
• Zero-frequency normalization inflates MTF values and consequently DQE.
• Simulated DQE inflation was 6% for a 10 cm ROI, increasing with smaller ROIs.
• Open-field normalization accurately determined MTF and NPS values.

Conclusions:

• Open-field normalization provides accurate MTF and DQE estimates.
• This method effectively avoids a common, non-obvious error inflating DQE by 5-20%.
• Adoption of open-field normalization is recommended for precise DQE calculations.