Diagnosis of Solid Breast Tumors with Sonographic Texture Analysis
Yu-Len Huang, Yaguang Liu · 2003
We evaluated a series of pathologically proven breast tumors using the support vector machine (SVM) in the differential diagnosis of solid breast tumors. This study evaluated 140 sonographical images of solid breast nodules (52 malignant and 88 benign). The physician located regions-of-interest (ROI) of ultrasonograhy and textual features were utilized to classify breast tumors. An SVM classifier using inter-pixel textual features classified the tumor as benign or malignant. The accuracy of SVM system for classifying malignancies is 95.0% (133/140), the sensitivity is 100% (52/52), the specificity is 92.0% (81/88), the positive predictive value is 88.1% (52/59), and the negative predictive value is 100% (81/81). The proposed computer-aided diagnostic system (CAD) differentiates solid breast nodules with a relatively high accuracy and helps inexperienced operators avoid misdiagnosis. The main advantage in the proposed system is that the training procedure of SVM was very fast and stable. The training and diagnosis procedure of the proposed SVM system is almost 700 times faster than that of multilayer perception neural networks (MLPs). With the growth of the database, new ultrasonic images can be collected and used as reference cases while performing diagnoses. This study reduces the training and diagnosis time dramatically.