Fig 1.
Flow chart of imaging data acquisition for dynamic contrast-enhanced digital mammogram.
Fig 2.
Grayscale bars with score finder above are an assumed grayscale representation of enhancement of suspicious lesion using a 10-point grayscale reference of iodine concentration within the breast lesion where pure white (score = 10) represents the highest level of contrast enhancement and black (score = 1) means poor contrast enhancement.
Fig 3.
Representative washout images of malignant breast lesion on DCE-mammogram of a 52-year-old woman with invasive breast cancer in right breast.
(a) Digital mammogram in MLO view before contrast agent injection shows focal asymmetry (arrows); (b-f) subtracted images at 2, 3, 4, 7 and 10 min after contrast agent injection. Subtracted images of breast malignancies (arrows) show early marked contrast enhancement and late-phase washout pattern.
Fig 4.
Representative persistent images of benign breast lesion on DCE-mammogram of a 45-year-old woman with fibroadenoma in left breast.
(a) Digital mammogram in MLO view before contrast agent injection shows a mass (arrow); (b-f) subtracted images at 2, 3, 4, 7 and 10 min after contrast agent injection. Subtracted images of fibroadenoma (arrow) show early moderate contrast enhancement and late-phase persistent pattern.
Table 1.
Characteristics of 148 breast lesions.
Fig 5.
DCE-mammogram demonstrating mean signal-intensity score of benign and malignant breast lesions from three readers and their average.
Malignant lesions show a significantly higher score than benign lesions on DCE-mammogram at all five time intervals (all p < 0.05) in three readers and their average. Figures present mean signal-intensity score and error bars present 95% CI. CI, confidence intervals.
Fig 6.
Areas under receiver operating characteristic curves (AUCs) for diagnostic performance of visual signal intensity (10-point scores) at five different time points determined by individual three readers and their average.
Table 2.
Multi-Reader Multi-Case (MRMC) performance analysis: Areas under receiver operating characteristic curves (AUCs) for enhancement score of each lesion by individual readers and average of three readers.
Table 3.
Sensitivity, specificity of best cut-off score for contrast enhancement to differentiate between benign and malignant breast lesions on DCE-mammogram at five different time intervals.
Table 4.
Kinetic patterns of DCE-mammogram between benign and malignant breast lesions using 2–4 and 2–10 min time intervals (reader C).