Generation of a suite of 3D computer-generated breast phantoms from a limited set of human subject data

Abstract

Purpose: The authors previously reported on a three-dimensional computer-generated breast phantom, based on empirical human image data, including a realistic finite-element based compression model that was capable of simulating multimodality imaging data. The computerized breast phantoms are a hybrid of two phantom generation techniques, combining empirical breast CT (bCT) data with flexible computer graphics techniques. However, to date, these phantoms have been based on single human subjects. In this paper, the authors report on a new method to generate multiple phantoms, simulating additional subjects from the limited set of original dedicated breast CT data. The authors developed an image morphing technique to construct new phantoms by gradually transitioning between two human subject datasets, with the potential to generate hundreds of additional pseudoindependent phantoms from the limited bCT cases. The authors conducted a preliminary subjective assessment with a limited number of observers (n = 4) to illustrate how realistic the simulated images generated with the pseudoindependent phantoms appeared.

Methods: Several mesh-based geometric transformations were developed to generate distorted breast datasets from the original human subject data. Segmented bCT data from two different human subjects were used as the "base" and "target" for morphing. Several combinations of transformations were applied to morph between the "base' and "target" datasets such as changing the breast shape, rotating the glandular data, and changing the distribution of the glandular tissue. Following the morphing, regions of skin and fat were assigned to the morphed dataset in order to appropriately assign mechanical properties during the compression simulation. The resulting morphed breast was compressed using a finite element algorithm and simulated mammograms were generated using techniques described previously. Sixty-two simulated mammograms, generated from morphing three human subject datasets, were used in a preliminary observer evaluation where four board certified breast radiologists with varying amounts of experience ranked the level of realism (from 1 = "fake" to 10 = "real") of the simulated images.

Results: The morphing technique was able to successfully generate new and unique morphed datasets from the original human subject data. The radiologists evaluated the realism of simulated mammograms generated from the morphed and unmorphed human subject datasets and scored the realism with an average ranking of 5.87 ± 1.99, confirming that overall the phantom image datasets appeared more "real" than "fake." Moreover, there was not a significant difference (p > 0.1) between the realism of the unmorphed datasets (6.0 ± 1.95) compared to the morphed datasets (5.86 ± 1.99). Three of the four observers had overall average rankings of 6.89 ± 0.89, 6.9 ± 1.24, 6.76 ± 1.22, whereas the fourth observer ranked them noticeably lower at 2.94 ± 0.7.

Conclusions: This work presents a technique that can be used to generate a suite of realistic computerized breast phantoms from a limited number of human subjects. This suite of flexible breast phantoms can be used for multimodality imaging research to provide a known truth while concurrently producing realistic simulated imaging data.

Figure 1

Radial grid patterns for mesh generation. This figure illustrates how the radial grids implement mapping from the “base” to the “target” in order to implement deformations. (a) “base” breast radial grid; (b) “target” breast radial grid (direct translation); (c) “target” breast radial grid (glandular data translation – stretch toward edge); (d) “target” breast radial grid (glandular data translation – shrink toward center); (e) “target” breast radial grid (nonuniform skewing).

Figure 2

Demonstration of the combination of glandular tissue from two different breasts. (a) Binary coronal slice from the “base” breast; (b) binary coronal slice from the “target’ breast; (c) distance transform image from the “base” breast; (d) distance transform image from the “target” breast; (e) smaller structures chosen from the “base” breast; (f) larger structures chosen from the “target” breast; (g) morphed breast glandular data resulting from the combination of the “base” and “target” breasts glandular data.

Figure 3

Sample coronal slices showing the effects of specific deformations. (a) Coronal slice from the “base” breast; (b) coronal slice from the “target” breast; (c) scaling the “base” breast glandular tissue to fit into the shell of the “target” breast; (d) scaling the “base” breast glandular tissue and fitting it into a resized shell of the “target” breast; (e) scaling and rotating the “base” breast glandular tissue and fitting it into the shell of the target breast; (f) scaling the “base” breast glandular tissue and fitting it into a rotated shell of the target breast; (g) scaling the “base” breast glandular tissue using the glandular data translation function to shrink the tissue toward the center of the “target” breast shell; (h) scaling and nonuniformly skewing the “base” breast glandular tissue and fitting it into the “target” breast shell; (i) scaling and eroding the “base” breast glandular tissue and fitting it into the “target” breast shell; (j) combining the “base” breast glandular tissue smaller structures with the “target” breast glandular tissue larger structures and fitting it into the “target” breast shell; and (k) combining the “target” breast glandular tissue smaller structures with the “base” breast glandular tissue larger structures and fitting it into the “target” breast shell.

Figure 4

Sample slices showing how combining different deformations affects the size, shape, and glandular distribution of the newly created morphed breast phantoms. (a) Coronal slice from the “base” breast; (b) coronal slice from the “target” breast; (c) combining the “base” breast glandular tissue smaller structures with the “target” breast glandular tissue larger structures and fitting it into a smaller “target” breast shell; (d) directly translating the “base” breast tissue into a resized smaller “target” breast shell; (e) directly translating the “base” breast tissue into the “target” breast shell; (f) rotating, nonlinearly skewing, and shrinking the “base” tissue toward the center of the “target” breast shell and combining with the larger structures of the “target” breast tissue; (g) rotating and shrinking the “base” breast tissue toward the center of the “target” breast shell and rotating the “target” breast shell; (h) rotating and nonlinearly skewing the “base” breast tissue, and then fitting it into a rotated and resized smaller “target” breast shell; (i) rotating and eroding the “base” breast glandular tissue and fitting it into a rotated and resized “target” breast shell; and (j) combining the “target” breast tissue smaller structures with the larger structures of the “base” breast and fitting it into a resized “base” breast shell.

Figure 5

Simulated CC mammograms ranging from realism rankings of 6–8. (a) 14% dense original breast – median score 6; (b) 28% dense original breast – median score 7.5; (c) 38% dense original breast – median score 6.5; (d) deformed/morphed (a) and (b) - key morphing parameters were shape morphing and glandular combination – median score 7.5; (e) deformed/morphed (a) and (b) - key morphing parameters were erosion and glandular combination – median score 8; (f) deformed/morphed (a) and (c) - key morphing parameters were 75° rotation, erosion, and glandular combination – median score 8; (g) deformed/morphed (b) and (c) - key morphing parameters were shape morphing and 60° rotation – median score 7; (h) deformed/morphed (a) and (b) - key morphing parameters were erosion, nonuniform skew, and 60° rotation – median score 6; (i) deformed/morphed (b) and (c) - key morphing parameters were rotation, erosion, and glandular combination – median score 6.5.