Accurate volume image reconstruction for digital breast tomosynthesis with directional-gradient and pixel sparsity...

Area of Science:

• Medical Imaging
• Radiology
• Image Reconstruction

Background:

• Accurate quantification of iodinated contrast agent (ICA) in contrast-enhanced digital breast tomosynthesis (DBT) is crucial for effective cancer diagnosis.
• Current DBT techniques face challenges in precise volumetric imaging of ICA distribution.

Purpose of the Study:

• To develop an accurate volumetric quantitative imaging method for ICA in contrast-enhanced DBT.
• To enhance the visualization and quantification of ICA for improved tumor detection and characterization.

Main Methods:

• Utilized dual-energy DBT (DE-DBT) scans.
• Developed an optimization-based image reconstruction algorithm incorporating sparsity regularization.
• Validated the algorithm using 2D/3D simulations and physical DE-DBT scans, including an anthropomorphic phantom.

Main Results:

• Demonstrated accurate and stable 2D image reconstruction.
• Showcased the algorithm's ability to resolve overlapping objects in 3D.
• Achieved accurate recovery of irregular ICA distribution in a tumor model and reduced depth-blurring in phantom studies.

Conclusions:

• The proposed sparsity regularization algorithm applied to DE-DBT shows significant promise for quantitative ICA imaging.
• Further research is needed to assess algorithm performance across diverse subject complexities for clinical application.