Spatial resolution analysis of computed tomographic images

Area of Science:

• Medical Imaging
• Radiology
• Image Analysis

Background:

• Accurate quantification of spatial resolution is crucial for interpreting medical images.
• Existing methods for Modulation Transfer Function (MTF) computation in X-ray computed tomography (CT) have limitations.
• Comparing model-dependent and model-independent approaches is essential for robust resolution assessment.

Purpose of the Study:

• To present and compare model-dependent and model-independent methods for quantifying spatial resolution in X-ray CT images.
• To evaluate the accuracy and applicability of these methods using phantom studies.
• To provide a framework for consistent MTF computation in CT imaging.

Main Methods:

• Development of model-dependent methods involving multiparameter fits of a 2D model function to image data.
• Application of model-independent methods through numerical transformation of image data.
• Utilizing phantom images with point, line, edge, and ring discontinuities for analysis.

Main Results:

• Comparison of MTF results obtained from both model-dependent and model-independent techniques.
• Presentation of spatial resolution measurements for two different CT scanners (Imatron C-100 and a UCSF experimental scanner).
• Validation of model-dependent predictions against model-independent results.

Conclusions:

• Both model-dependent and model-independent methods can effectively quantify spatial resolution in CT images.
• The study provides valuable data for understanding and comparing different resolution quantification techniques.
• Results contribute to the standardization and improvement of CT image quality assessment.