[Characterization of sonographically detected breast lesions using three-dimensional data sets]

Area of Science:

• Radiology and Imaging Science
• Oncology and Breast Cancer Diagnostics

Background:

• Accurate characterization of breast lesions is crucial for timely diagnosis and treatment of breast cancer.
• Traditional 2D ultrasound (US) is a standard imaging modality, but advanced techniques like 3D US offer potential improvements in lesion assessment.

Purpose of the Study:

• To evaluate the diagnostic performance of 3D ultrasound compared to 2D ultrasound for breast lesions.
• To correlate ultrasound imaging criteria with histological findings and rank their significance in lesion characterization.

Main Methods:

• Ninety-two breast lesions (61 malignant, 31 benign) were examined using 2D and 3D ultrasound.
• Histological confirmation via core biopsy was performed.
• Lesions were assessed using BI-RADS classification, with statistical analysis including Pearson's chi-square and kappa statistics for comparison.

Main Results:

• 3D ultrasound demonstrated diagnostic value comparable to 2D ultrasound, with minor differences in specific measurements.
• Significant 2D criteria for malignancy included lesion margin and effect on adjacent tissue; 3D US highlighted shape, margin in multiple views, and frontal plane characteristics.
• Sensitivities and specificities were similar between 2D and 3D US, while 3D US provided enhanced visualization of the vascular tree.

Conclusions:

• 3D ultrasound is a valuable tool for breast cancer diagnosis, performing comparably to 2D ultrasound.
• The evaluation of frontal views in 3D imaging is significant for accurate lesion characterization.