Combining low-, high-level and empirical domain knowledge for automated segmentation of ultrasonic breast lesions.

Breast cancer is the most frequently diagnosed malignancy and the second leading cause of mortality in women . In the last decade, ultrasound along with digital mammography has come to be regarded as the gold standard for breast cancer diagnosis. Automatically detecting tumors and extracting lesion...

Descrizione completa

Dettagli Bibliografici
Pubblicato in:	IEEE Transactions on medical imaging 22, 2 (2003).
Autore principale:	Madabhushi, A.
Natura:	Articolo
Lingua:	English
Soggetti:	1.8 GHz. 1.8-GHz Pentium machine. 18 s. Stavros Criteria. Algorithms. Automated segmentation. Average mean boundary error. Breast cancer diagnosis. Deformable model. Deformable shape-based model. Digital mammography. Directional gradient. Empirical domain knowledge. Empirical domain specific knowledge. Empirical rules. High-level domain knowledge. Image pixels. Intensity. Lesion boundaries. Low-level domain knowledge. Mathematical formulation. Normalized true positive area overlap. Probabilistic classification. Recursive refinement. Running time. Seed point. Shadowing artifacts. Similar tumor like structures. Sonographic lesions. Speckle noise. Texture. Training samples. Tumor boundaries. Tumor regions. Tumors. Ultrasonic breast lesions.