Optimization methods for very accurate Digital Breast Tomosynthesis image reconstruction

TL;DR

This paper introduces highly accurate, convergent iterative optimization algorithms for Digital Breast Tomosynthesis image reconstruction, improving early detection of cancerous features while aligning with clinical needs.

Contribution

It proposes a novel optimization framework with Total Variation regularization and automatic parameter tuning for improved breast image reconstruction.

Findings

Effective detection of masses and microcalcifications in early iterations

Enhanced image quality with prolonged algorithm execution

Validated on real phantom and clinical data

Abstract

Digital Breast Tomosynthesis is an X-ray imaging technique that allows a volumetric reconstruction of the breast, from a small number of low-dose two-dimensional projections. Although it is already used in clinical setting, enhancing the quality of the recovered images is still a subject of research. Aim of this paper is to propose, in a general optimization framework, very accurate iterative algorithms for Digital Breast Tomosynthesis image reconstruction, characterized by a convergent behaviour. They are able to detect the cancer object of interest, i.e. masses and microcalcifications, in the early iterations and to enhance the image quality in a prolonged execution. The suggested model-based implementations are specifically aligned to Digital Breast Tomosynthesis clinical requirements and take advantage of a Total Variation regularizer. We also tune a fully-automatic strategy to set a proper regularization parameter. We assess our proposals on real data, acquired from a breast accreditation phantom and a clinical case. The results confirm the effectiveness of the presented solutions in reconstructing breast volumes with particular focus on the masses and microcalcifications.