A virtual imaging study of microcalcification detection performance in digital breast tomosynthesis: Patients versus 3D textured phantoms

Abstract

Background: Clinical studies to evaluate the performance of new imaging devices require the collection of patient data. Virtual methods present a potential alternative in which patient-simulating phantoms are used instead.

Purpose: This work uses a virtual imaging technique to examine the extent to which human observer microcalcification detection performance in phantom backgrounds matches that in real patient backgrounds for digital breast tomosynthesis (DBT).

Methods: This work used the following DBT image datasets: (1) 142 real patient images and (2) 20 real images of the physical L1 phantom, both acquired on a GEHC Senographe Pristina system; (3) 217 simulated images of the Stochastic Solid Breast Texture (SSBT) phantom and (4) 217 simulated images of the digital L1 phantom, both created with the CatSim framework. The L1 phantom is a PMMA container filled with water and PMMA spheres of varying diameters. The SSBT phantom is a computational phantom composed of glandular and adipose tissue compartments. Signal-present images were generated by inserting simulated microcalcification clusters, containing individual calcifications with thicknesses and projected areas in the range of 165-180 µm, 195-210 µm and 225-240 µm, and 0.025-0.031 mm², 0.032-0.040 mm², 0.041-0.045 mm² respectively, at random locations into all four background types. Three human observers performed a search/localization task on 120 signal-present and 97 signal-absent volumes of interest (VOIs) per background type. A jackknife alternative free-response receiver operating characteristic (JAFROC) analysis was applied to calculate the area under the curve (AUC). The simulation procedure was first validated by testing the physical and digital L1 background AUC values for equivalence (margin = 0.1). The AUC for patient backgrounds and each phantom type (SSBT, physical L1, digital L1) was then compared. Additionally, each patient's VOI was categorized in homogeneous or heterogeneous background texture distribution by an experienced physicist, and by local volumetric breast density (VBD) at the insertion position to examine their effect on correctly detected fraction of microcalcification clusters.

Results: Mean AUC for the patient images was 0.70 ± 0.04, while mean AUCs of 0.74 ± 0.04, 0.76 ± 0.03, and 0.76 ± 0.07 were found for the SSBT, physical L1 and digital L1 phantoms, respectively. The AUC for the physical and digital L1 phantoms was equivalent (p = 0.03), as well as for the patients and SSBT backgrounds (p = 0.002). The physical and digital L1 images did not have equivalent detection performance compared to patient images (p = 0.06 and p = 0.9, respectively). In patient backgrounds, the correctly detected fraction of microcalcifications clusters fell from 0.53 for the lowest density (VBD < 4.5%) to 0.40 for the highest density (VBD ≥ 15.5%). Microcalcification detection fractions were 0.52, 0.55, and 0.55 for the SSBT, physical L1 and digital L1 backgrounds, respectively.

Conclusions: Detection levels were equivalent between the physical and digital versions of the L1 phantom. Detection in L1 and patient backgrounds was not equivalent, however, differences in detection performance were small, confirming the potential value of this phantom. The digital SSBT phantom was found to be equivalent to patient backgrounds for DBT studies of microcalcification cluster detection performance, for the DBT system and reconstruction algorithm used in this study.

Keywords: detection; digital breast tomosynthesis; digital phantoms; virtual imaging trial.

FIGURE 1

Central reconstructed DBT slice (40 mm × 40 mm) of three SSBT phantoms generated with different texture configurations.

FIGURE 2

Physical L1 phantom (a), along with central reconstructed DBT slice (40 mm × 40 mm) of the physical L1 phantom (b) and the digital L1 phantom (c).

FIGURE 3

In‐plane reconstructed DBT images of two microcalcification clusters simulated in a patient image, an SSBT phantom, and a physical and digital L1 phantom.

FIGURE 4

Box plot of the mean signal intensity measured in a 40 mm × 40 mm ROI in the 0° DBT projection images. The lines out from the box indicate the maximum and minimum data values, and the bottom and top of the box mark the 25th and 75th percentiles of the mean values. The line inside the box indicates the median, and the marker the mean.

FIGURE 5

Power spectra calculated from the 0° projection from each scan, for the four backgrounds. Each solid line shows the average value for a given background, while the shaded regions show the 5% and 95% extent of the power spectra for a given background. (a) The four backgrounds compared; (b) SSBT (simulated) versus patients (real); (c) Digital L1 (simulated) versus patients (real); and (d) Physical L1 (real) versus patients (real).

FIGURE 6

Average power spectrum value S_avg at 1, 1.5, and 2 mm⁻¹ calculated from the 0° projection images for each background type. The error bar indicates the coefficient of variation.

FIGURE 7

The reader‐averaged alternative FROC curves for the four different background types. FPF, false positive fraction; LLF, lesion localization fraction.

FIGURE 8

Average correctly detected fraction of the clusters simulated in the four backgrounds. The patient backgrounds are subdivided by local breast density, group 1 (VBD < 4.5%), group 2 (4.5 % ≤ VBD < 7.5%), group 3 (7.5 % ≤ VBD < 15.5%), and group 4 (VBD ≥ 15.5%) (left) and visual assessment of homogeneous versus heterogeneous background (right). The error bars indicate the 95% confidence interval.