Simulation of high-resolution test objects using non-isocentric acquisition geometries in next-generation digital...

Area of Science:

• Medical Imaging
• Radiology
• Image Reconstruction

Background:

• Conventional digital breast tomosynthesis (DBT) uses isocentric acquisition geometry, leading to imaging artifacts that degrade image quality.
• Artifacts like aliasing are inherent to standard DBT, limiting the diagnostic accuracy of reconstructed images.
• The next-generation tomosynthesis (NGT) prototype was developed to explore alternative acquisition geometries for DBT.

Purpose of the Study:

• To investigate non-isocentric acquisition geometries for DBT using the NGT prototype.
• To compare the high-frequency performance and multiplanar reconstruction (MPR) quality of non-isocentric versus isocentric DBT acquisition.
• To develop custom acquisition geometries for DBT using virtual clinical trials (VCTs).

Main Methods:

• Simulated virtual clinical trials (VCTs) using high-resolution star and tilted bar pattern test objects.
• Implemented two source- and detector-motion paths for non-isocentric geometries to achieve super-sampled reconstructions.
• Evaluated spatial resolution using an aliasing-sensitive metric and compared MPR quality between isocentric and non-isocentric acquisitions.

Main Results:

• Non-isocentric acquisition geometries demonstrated significant improvement over isocentric geometries, with up to 75.2% enhancement in spatial resolution.
• The greatest improvement was observed for frequencies perpendicular to the x-ray source motion, mitigating DBT's aliasing susceptibility.
• Non-isocentric reconstructions exhibited super-resolution in both frequency orientations and superior MPR quality without z-dependent degeneracies.

Conclusions:

• Non-isocentric acquisition geometries offer a substantial advancement over conventional isocentric methods for DBT.
• These novel geometries effectively reduce artifacts and enhance spatial resolution, leading to improved image quality.
• The findings support the clinical utility of non-isocentric DBT for more accurate breast cancer detection.