(Tuesday, 7/31/2018) 1:15 PM - 1:45 PM
Room: Exhibit Hall | Forum 1
Purpose: We have conducted virtual clinical trials (VCTs) of breast lesion detection in digital breast tomosynthesis (DBT), using OpenVCT a novel open-source simulation pipeline, to test in-silico the factors which influence the detectability of microcalcifications. In previous work, our VCTs showed an average AUC improvement for DBT vs DM of 0.027 for microcalcifications and 0.103 for masses, in close agreement (within 1%) of clinical data reported in the literature.
Methods: For this test, we generated 12 software breast phantoms (volume 700ml, compressed thickness 6.33cm), varying the number of simulated tissue compartments and their shape. Into each phantom, we inserted multiple lesions located 1.5 cm apart in the plane parallel to detector at the level of the nipple. Simulated single calcifications of various size and composition were inserted. DBT projections of phantoms with and without lesions were synthesized assuming a clinical acquisition geometry and automatic exposure control. Reconstructed DBT slices were produced using a commercially available software library. Lesion detection was simulated with channelized Hotelling observers, having 15 LG channels and a spread of 22, using independent training and test sets.
Results: The following factors were tested: source motion (i.e., step-and-shoot vs. continuous tube motion), detector element size, reconstructed voxel size, and super-resolution reconstructions. Preliminary results indicate that each of these factors can affect calcification detectability, especially for small calcifications. Those factors which reduce blurring of the calcifications in the projections images can likewise reduce blurring in the reconstructed images, thereby improving detectability.
Conclusion: In VCT of DBT, the detection of small calcifications is affected by source motion, detector element size, reconstructed voxel size, and the use of super-resolution reconstructions.
Funding Support, Disclosures, and Conflict of Interest: ADM is a member of the scientific advisory board and a shareholder of Real-Time Tomography, LLC
Keywords
Breast, Tomosynthesis, Calcifications
Taxonomy
IM- Breast x-ray Imaging: Digital Breast Tomosynthesis (DBT)
Contact Email
