Assessment of Volumetric Dense Tissue Segmentation in Tomosynthesis Using Deep Virtual Clinical Trials.

The adoption of artificial intelligence (AI) in medical imaging requires careful evaluation of machine-learning algorithms. We propose the use of a "deep virtual clinical trial" (DeepVCT) method to effectively evaluate the performance of AI algorithms. In this paper, DeepVCTs have been proposed to elucidate limitations of AI applications and predictions of clinical outcomes, avoiding biases in study designs. The DeepVCT method was used to evaluate the performance of nnU-Net models in assessing volumetric breast density (VBD) from digital breast tomosynthesis (DBT) images. In total, 2,010 anatomical breast models were simulated. Projections were simulated using the acquisition geometry of a clinical DBT system. The projections were reconstructed using 0.1, 0.2, and 0.5 mm plane spacing. nnU-Net models were developed using the center-most planes of the reconstructions with the respective ground-truth. The results show that the accuracy of the nnU-Net improves significantly with DBT images reconstructed with 0.1 mm plane spacing (78.4×205.3×40.1 mm3). The segmentations resulted in Dice values up to 0.84 with area under the receiver operating characteristic curve of 0.92. The optimization of plane spacing for VBD assessment was used as an exemplar of a DeepVCT application, allowing us to interpret better the input parameters and outcomes of the nnU-Net. Thus, DeepVCTs can provide a plethora of evidence to predict the efficacy of these algorithms using large-scale simulation-based data.

Comparison of radiation dose between 2D digital stereotactic versus digital breast tomosynthesis-guided breast biopsies.

RATIONALE AND OBJECTIVE: Use of digital breast tomosynthesis (DBT) in breast imaging has necessitated DBT-guided biopsy, however, a single DBT acquisition may result in a greater radiation dose than a single DM acquisition. Our objective was to compare the number of images acquired and the resulting radiation dose of DBT versus DM-guided breast biopsies. METHOD: All biopsies performed on our DM unit from 8/2016 to 1/2017 and on our DM-DBT unit from 8/2017 to 1/2018 were retrospectively reviewed. The number of image acquisitions, average glandular dose (AGD) per acquisition and per procedure were computed and stratified by guidance modality and lesion type. RESULTS: 25 DM-guided biopsies were performed on the DM-only unit, 58 biopsies were performed with DM guidance on the dual unit (DM-DU) and 29 were performed with DBT. The average number of images acquisitions was 10.9 for DM-only unit biopsies, 9.3 images for DM-DU biopsies and 4.3 images for DBT-guided biopsies. Mean procedure AGD for DM-only unit biopsies was 28.77 mGy, versus 22.06 mGy for DM-DU and 10.18 mGy for DBT biopsies. Mean procedure AGD for biopsied calcification-only lesions was 22.3 mGy for DM-DU versus 10.7 mGy for DBT guidance (p < 0.001), with an average of 8.1 images per procedure for DM-DU versus 4.2 for DBT. CONCLUSION: Fewer image acquisitions were obtained with DBT compared with DM guidance, therefore, the overall dose of DBT-guided procedures was less. The dose reduction obtained with DBT is possible across all lesion types, even for calcification-only lesions.