Digital Breast Tomosynthesis: Towards Dose Reduction through Image Quality Improvement

Abstract

Currently, breast cancer is the most commonly diagnosed type of cancer worldwide. Digital Breast Tomosynthesis (DBT) has been widely accepted as a stand-alone modality to replace Digital Mammography, particularly in denser breasts. However, the image quality improvement provided by DBT is accompanied by an increase in the radiation dose for the patient. Here, a method based on 2D Total Variation (2D TV) minimization to improve image quality without the need to increase the dose was proposed. Two phantoms were used to acquire data at different dose ranges (0.88-2.19 mGy for Gammex 156 and 0.65-1.71 mGy for our phantom). A 2D TV minimization filter was applied to the data, and the image quality was assessed through contrast-to-noise ratio (CNR) and the detectability index of lesions before and after filtering. The results showed a decrease in 2D TV values after filtering, with variations of up to 31%, increasing image quality. The increase in CNR values after filtering showed that it is possible to use lower doses (-26%, on average) without compromising on image quality. The detectability index had substantial increases (up to 14%), especially in smaller lesions. So, not only did the proposed approach allow for the enhancement of image quality without increasing the dose, but it also improved the chances of detecting small lesions that could be overlooked.

Keywords: digital breast tomosynthesis; dose reduction; image quality; total variation minimization.

Figure 1

Acrylic phantom simulating breast tissue and lesions of high attenuation (aluminum disks of different diameters and 1 mm thickness). Diameters of the disks of the first column (top to bottom): 5.0 mm, 4.0 mm, 3.0 mm, 2.0 mm, 1.0 mm, and 0.5 mm, respectively; Diameter of the disks of the second column (top to bottom): 4.0 mm, 2.0 mm, 0.5 mm, 1.0 mm, 3.0 mm, and 5.0 mm, respectively.

Figure 2

ROIs used to calculate CNR using the Gammex 156 phantom. Circular ROI over the 2 mm lesion-like mass and square background ROI centered in the lesion, excluding all voxels corresponding to the lesion.

Figure 3

(a) ROI used to compute the task-based transfer function (TTF). (b) ROIs used for the noise power spectrum (NPS) assessment.

Figure 4

The top row shows the synthesized signals of four different sizes to be detected with a designer contrast profile, and the bottom row, with a rectangular contrast profile. The Fourier transform of such a signal is the task function, $W_{task}$.

Figure 5

Values of CNR obtained for the original data and the filtered data of the Gammex 156 phantom as a function of dose. The purple arrows represents the possible dose reduction made by applying the filter to obtain the same CNR.

Figure 6

Images of a 2.0 mm mass of the Gammex 156 phantom obtained for the original (up row) and filtered data (down row) for each acquisition dose.

Figure 7

Images of a cluster of microcalcifications of the Gammex 156 phantom obtained for the original (up row) and filtered data (down row) for each acquisition dose.

Figure 8

Detectability index values obtained for a circular signal with designer profile and diameters of (a) 5 mm, (b) 3 mm, (c) 1 mm, and (d) 0.5 mm for each acquisition dose of original and filtered data. The purple arrows represents the possible dose reduction made by applying the filter to obtain the same detectability index.

Figure 9

Detectability index values obtained for a circular signal with rectangular profile and diameters of (a) 5 mm, (b) 3 mm, (c) 1 mm, and (d) 0.5 mm for each acquisition dose of original and filtered data. The purple arrows represent possible dose reduction made by applying the filter to obtain the same detectability index.