Utility of U-Net for the objective segmentation of the fibroglandular tissue region on clinical digital mammograms.

This study investigates the equivalence or compatibility between U-Net and visual segmentations of fibroglandular tissue regions by mammography experts for calculating the breast density and mean glandular dose (MGD). A total of 703 mediolateral oblique-view mammograms were used for segmentation. Two region types were set as the ground truth (determined visually): (1) one type included only the region where fibroglandular tissue was identifiable (called the 'dense region'); (2) the other type included the region where the fibroglandular tissue may have existed in the past, provided that apparent adipose-only parts, such as the retromammary space, are excluded (the 'diffuse region'). U-Net was trained to segment the fibroglandular tissue region with an adaptive moment estimation optimiser, five-fold cross-validated with 400 training and 100 validation mammograms, and tested with 203 mammograms. The breast density and MGD were calculated using the van Engeland and Dance formulas, respectively, and compared between U-Net and the ground truth with the Dice similarity coefficient and Bland-Altman analysis. Dice similarity coefficients between U-Net and the ground truth were 0.895 and 0.939 for the dense and diffuse regions, respectively. In the Bland-Altman analysis, no proportional or fixed errors were discovered in either the dense or diffuse region for breast density, whereas a slight proportional error was discovered in both regions for the MGD (the slopes of the regression lines were -0.0299 and -0.0443 for the dense and diffuse regions, respectively). Consequently, the U-Net and ground truth were deemed equivalent (interchangeable) for breast density and compatible (interchangeable following four simple arithmetic operations) for MGD. U-Net-based segmentation of the fibroglandular tissue region was satisfactory for both regions, providing reliable segmentation for breast density and MGD calculations. U-Net will be useful in developing a reliable individualised screening-mammography programme, instead of relying on the visual judgement of mammography experts.

Moderate versus high concentration of contrast material for aortic and hepatic enhancement and tumor-to-liver contrast at multi-detector row CT.

PURPOSE: To prospectively evaluate aortic and hepatic enhancement and depiction of hypervascular hepatocellular carcinoma (HCC) between two contrast materials with moderate and high iodine concentrations when administered at same iodine dose and injection duration at multi-detector row helical computed tomography (CT). MATERIALS AND METHODS: Institutional review board approval and informed patient consent were obtained. One hundred eighty-six patients were studied, and 67 patients with hypervascular HCC were identified. Ninety-four patients were assigned to receive iohexol 350 (mg iodine per milliliter) with protocol A; 92, iohexol 300 with protocol B. In both protocols, iohexol with same iodine load per weight (518 mg/kg) was administered with same injection duration (25 seconds). Multiphase CT scanning was started 10, 20, 50, and 180 seconds after the trigger (threshold level set at increase of 100 HU over baseline CT number of aorta). Enhancement of aorta and liver was measured in 186 patients. Tumor-to-liver contrast was measured in 67 patients with hypervascular HCC. Statistical analysis was performed with Mann-Whitney U test. RESULTS: Medians of aortic enhancement during four phases were 325, 185, 112, and 69 HU with protocol A. Corresponding values were 344, 266, 121, and 73 HU with protocol B. During all phases, aortic enhancement was significantly higher with protocol B (P = .046, P < .001, P < .001, and P = .002). Hepatic enhancement during four phases was 6, 21, 48, and 34 HU with protocol A. Corresponding values were 3, 17, 47, and 35 HU with protocol B. Hepatic enhancement was significantly higher with protocol A during first and second phases (P < .001 for both), although there was no significant difference between protocols during third and fourth phases (P = .778 and P = .178, respectively). Medians of tumor-to-liver contrast during four phases were 22, 34, 0.5, and -1.1 HU with protocol A. Corresponding values were 23, 45, 0, and -8.6 HU with protocol B. Tumor-to-liver contrast was significantly higher with protocol B during second phase (P = .049), although there was no difference between protocols during other phases. CONCLUSION: When total iodine dose was adjusted to body weight and injection duration was fixed, rapid administration of moderate concentration of contrast material was more effective for depiction of hypervascular HCC than was high concentration of contrast material.