Quantitative vascularity of breast masses by Doppler imaging: Regional variations and diagnostic implications

Citation
Cm. Sehgal et al., Quantitative vascularity of breast masses by Doppler imaging: Regional variations and diagnostic implications, J ULTR MED, 19(7), 2000, pp. 427-440
Citations number
24
Categorie Soggetti
Radiology ,Nuclear Medicine & Imaging
Journal title
JOURNAL OF ULTRASOUND IN MEDICINE
ISSN journal
02784297 → ACNP
Volume
19
Issue
7
Year of publication
2000
Pages
427 - 440
Database
ISI
SICI code
0278-4297(200007)19:7<427:QVOBMB>2.0.ZU;2-N
Abstract
Seventy-four biopsy proven breast masses were imaged by color and power Dop pler imaging to evaluate vascular pattern of malignant and benign breast ma sses. The images were analyzed for vascularity. The measurements were made over the entire mass as well as regionally at its core, at its periphery, a nd in the tissue surrounding it. The surgical specimens were analyzed for m icrovessel density. The diagnostic performance of Doppler sonographic vascu larity indices was evaluated by receiver operating characteristic analysis. The malignant masses were 14 to 54% more vascular than the benign masses. Both types of masses were more vascular by ultrasonography than the tissue surrounding them. Whereas benign masses were 2.2 times more vascular than t he surrounding tissue, the malignant masses were 5.0 times more vascular. I n a subset of patients the regional vascularity at the core, periphery, and surrounding tissue by Doppler imaging exhibited a strong correlation (R-2 > 0.9) with the corresponding histologic microvessel density measurements. Although the malignant masses exhibited a strong gradient in vascularity, c ore, periphery, surrounding tissue, the benign masses had relatively unifor m distribution of vascularity. The area under the receiver operating charac teristic curve (A(z)) for the Doppler indices ranged from 0.56 +/- 0.07 to 0.65 +/- 0.07. A nonlinear analysis including age-specific values of Dopple r indices improved the diagnostic performance to A(z) = 0.85 +/- 0.06. In c onclusion, quantitative Doppler imaging when used in combination with a non linear rule-based approach has the potential for differentiating between ma lignant and benign masses.