A Sphere Phantom Approach to Measure Directional Modulation Transfer Functions for Tomosynthesis Imaging Systems

Area of Science:

• Medical Imaging Physics
• Tomosynthesis System Characterization
• Image Quality Assessment

Background:

• Tomosynthesis imaging systems suffer from artifacts, complicating accurate image quality assessment.
• Traditional methods for isolating phantom data are insufficient due to these artifacts.
• Measuring spatially varying directional modulation transfer functions (MTFs) is crucial for tomosynthesis system evaluation.

Purpose of the Study:

• To introduce a novel sphere phantom approach for measuring directional MTFs in tomosynthesis.
• To develop an optimized background detrending technique for accurate sphere-only data acquisition.
• To reliably quantify tomographic imaging system performance across various directions.

Main Methods:

• A sphere phantom was utilized to acquire data for tomosynthesis imaging.
• An optimized background detrending technique was applied to local volumes with varying cone angles.
• Directional plane spread functions (PlSFs) were estimated using Richardson-Lucy deconvolution with Tikhonov-Miller to mitigate noise.
• Directional MTFs were computed via the Fourier transform of the estimated PlSFs.

Main Results:

• The proposed method successfully measured directional MTFs along arbitrary directions.
• The technique demonstrated reliability, especially in characterizing low-frequency performance.
• Measured directional MTFs showed good agreement with ideal MTFs derived from simulated point objects.

Conclusions:

• The sphere phantom approach provides a robust method for directional MTF measurement in tomosynthesis.
• The optimized detrending technique effectively addresses artifacts in tomosynthesis reconstruction.
• This method enhances the ability to assess and compare tomosynthesis imaging system performance.