Feature Extraction from a Signature Based on the Turning Angle Function for the Classification of Breast Tumors

Abstract

Malignant breast tumors and benign masses appear in mammograms with different shape characteristics: the former usually have rough, spiculated, or microlobulated contours, whereas the latter commonly have smooth, round, oval, or macrolobulated contours. Features that characterize shape roughness and complexity can assist in distinguishing between malignant tumors and benign masses. Signatures of contours may be used to analyze their shapes. We propose to use a signature based on the turning angle function of contours of breast masses to derive features that capture the characteristics of shape roughness as described above. We propose methods to derive an index of the presence of convex regions (XR ( TA )), an index of the presence of concave regions (VR ( TA )), an index of convexity (CX ( TA )), and two measures of fractal dimension (FD ( TA ) and FDd ( TA )) from the turning angle function. The methods were tested with a set of 111 contours of 65 benign masses and 46 malignant tumors with different parameters. The best classification accuracies in discriminating between benign masses and malignant tumors, obtained for XR ( TA ), VR ( TA ), CX ( TA ), FD ( TA ), and FDd ( TA ) in terms of the area under the receiver operating characteristics curve, were 0.92, 0.92, 0.93, 0.93, and, 0.92, respectively.

Assessing the performance of morphological parameters in distinguishing breast tumors on ultrasound images

“…Another well-established parameter is the circularity [11,15,18], a contour descriptor that proved to be significant in the classification of breast tumors [19]. Furthermore, techniques such as normalized radial length [11], normalized radial gradient [13,14], convex polygon [16,18,20], fractal analysis [21] and signature of contours [22] have been used to quantify tumor contour irregularities.…”

Effective Shape Measures in Malignant Risk Assessment for Breast Tumor on Sonography

“…To compare the performance of individual convexity measure, two shape measures used to quantify the irregularity, roughness, and spiculation index of breast tumor including 2D box-counting fractal dimensions (FD) and CX TA [11] were also computed from the segmented contours. The statistics of these measures are summarized in Table 1.…”

“…The researchers have developed various measures of shape feature for breast lesions on 2D sonography [6][7][8][9], such as compactness, lobulation index [6], elliptic-normalized circumference (ENC) [6], spiculation measure [7], convex hull depth [8], etc. Other similar objective shape measures, such as spiculation index (SI TF ) [10] and index of convexity (CX TA ) [11] from turning angle function have been also developed to distinguish benign masses and malignant tumors on X-ray mammography.…”

Image classification: A novel texture signature approach

“…Rubner et al used signature technique to perform texture-based image retrieval [4]. Guliato et al [5] used a signature based on turning angle function of contour of breast masses to encode features that characterise contours roughness and complexity for breast tumour classification. Zwiggelaar et al [6] investigated scale-orientation signature for labeling of structures in images and to classify pixels into linear structures, blob-like structures or background texture.…”