Simulation of 3D DLA masses in digital breast tomosynthesis

Abstract

Digital breast tomosynthesis (DBT) is suggested to have superior performance compared to 2D mammography in terms of cancer visibility, especially in the case of dense breasts. However, the overall performance of tomosynthesis for screening applications, and the manner in which tomosynthesis should be optimally used for screening remains unclear. This motivates the development of software tools that can insert user-defined synthetic pathology of realistic appearance into clinical tomosynthesis images for subsequent use in virtual clinical trials. We present a method for inserting lesions grown using Diffusion Limited Aggregation, previously validated in 2D mammograms, into clinical DBT images. A preliminary pilot study was used to validate the realism of the masses, wherein three readers each viewed 19 cases and rated the realism of the inserted masses. Each case included a simulated mass inserted in the tomosynthesis projections and the counterpart digital 2D mammogram. These results show that masses can be successfully embedded in the tomosynthesis projections and can produce visually authentic DBT images containing synthetic pathology. These results will be used to further optimize the appearance of these masses in DBT for an upcoming validation.

Design and validation of biologically inspired spiculated breast lesion models utilizing structural tissue distortion

“…Inserting these into a variety of different clinical tissue backgrounds using the aforementioned blending process with template multiplication 3,4 has produced visually realistic results as shown in Figure 10. Figure 7 also shows a standard DLA mass subject to the distortion process and inserted into a clinical image.…”

“…Therefore, the internal structure of the spicule horn manifolds were filled with hair-like strands with each hair given the density of glandular tissue, so that spicule density could be controlled by varying the packing fraction of the resulting strands. Collections of spicules were then attached to solid lesion cores created using diffusion limited aggregation (DLA) 3,4 as shown in Figure 1e-f. …”

“…For the extended observer study, a total of 43 spiculated lesions were simulated and inserted into real clinical images again using template multiplication 3,4 . The feedback from the pilot study was used to produce more realistic looking spiculated lesions.…”

“…For the pre-pilot study, eleven lesion models were inserted into a mixture of 2D and DBT image data acquired in the Hologic system (Hologic Inc, Massachusetts, USA). The simulated spiculated lesions were inserted into real clinical images using template multiplication 3,4 . The observer was then asked to rate these on a 3-point scale as either 'real', 'equivocal', or 'simulated'.…”

“…A(r The OPTIMAM image simulation toolbox 2 is a complete end-to-end VCT pipeline that can generate synthetic masses 3,4 or calcifications 5 and insert these into unprocessed clinical digital mammography images or hybrid models of breast tissue 6 . These are then processed using an image acquisition tool that can currently simulate either the Hologic Selenia 2 or the Siemens MAMMOMAT Inspiration 7 imaging systems in 2D and tomosynthesis operating modes.…”