Modeling the Dermoscopic Structure Pigment Network Using a Clinically Inspired Feature Set

IEEE Trans. Med. Imaging

Lee

McLean

et al. 2013

112

Abstract-Irregular streaks are important clues for Melanoma (a potentially fatal form of skin cancer) diagnosis using dermoscopy images. This paper extends our previous algorithm to identify the absence or presence of streaks in a skin lesions, by further analyzing the appearance of detected streak lines, and performing a three-way classi cation for streaks, Absent, Regular, and Irregular, in a pigmented skin lesion. In addition, the directional pattern of detected lines is analyzed to extract their orientation features in order to detect the underlying pattern. The method uses a graphical representation to model the geometric pattern of valid streaks and the distribution and coverage of the structure. Using these proposed features of the valid streaks along with the color and texture features of the entire lesion, an accuracy of 76.1% and weighted average area under ROC curve (AUC) of 85% is achieved for classifying dermoscopy images into streaks Absent, Regular, or Irregular on 945 images compiled from atlases and the internet without any exclusion criteria. This challenging dataset is the largest validation dataset for streaks detection and classi cation published to date. The data set has also been applied to the two-class sub-problems of Absent/Present classi cation (accuracy of 78.3% with AUC of 83.2%) and to Regular/Irregular classi cation (accuracy 83.6% with AUC of 88.9%). When the method was tested on a cleaned subset of 300 images randomly selected from the 945 images, the AUC increased to 91.8%, 93.2% and 90.9% for the Absent/Regular/Irregular, Absent/Present, and Regular/Irregular problems, respectively.

Section: A Identifying Valid Streak Lines From Candidate Streaksmentioning

confidence: 99%

Detection and Analysis of Irregular Streaks in Dermoscopic Images of Skin Lesions

IEEE Trans. Med. Imaging

Lee

McLean

et al. 2013

112

“…As colour features, we generate mean; standard deviation; the ratio of these; and entropy of each channel, in addition to |var(η 1 ) − var(η 2 )|, adding up to a 13-D colour feature vector. Further, we add texture features to our colour feature-vectors, in a similar fashion as in [10]: four of the classical statistical texture measures of [11] (contrast, correlation, homogeneity and energy) are derived from the grey level co-occurrence matrix (GLCM) of each channel. This is an additional 12-D texture feature vector; thus we arrive at a 25-D feature vector.…”

Section: Texture and Colour Feature Vectorsmentioning

confidence: 99%

Intrinsic Melanin and Hemoglobin Colour Components for Skin Lesion Malignancy Detection

Madooei

Drew

Lecture Notes in Computer Science

et al. 2012

Self Cite

Abstract. In this paper we propose a new log-chromaticity 2-D colour space, an extension of previous approaches, which succeeds in removing confounding factors from dermoscopic images: (i) the effects of the particular camera characteristics for the camera system used in forming RGB images; (ii) the colour of the light used in the dermoscope; (iii) shading induced by imaging non-flat skin surfaces; (iv) and light intensity, removing the effect of light-intensity falloff toward the edges of the dermoscopic image. In the context of a blind source separation of the underlying colour, we arrive at intrinsic melanin and hemoglobin images, whose properties are then used in supervised learning to achieve excellent malignant vs. benign skin lesion classification. In addition, we propose using the geometric-mean of colour for skin lesion segmentation based on simple greylevel thresholding, with results outperforming the state of the art.

“…As a fundamental step towards computer-aided diagnosis of skin cancers, automatic detection of many of these dermoscopic structures have been recently addressed in the literature [3,4]. However, the automatic detection of streaks has only recently been investigated [5,6].…”

Section: Mathematical Definitionmentioning

confidence: 99%

“…In STR, Structural set includes the number of candidate streaks in the image, average number of pixels of candidate streaks, and the ratio of the streaks size to the lesion size in pixels; Chromatic set consists of the mean, standard deviation and reciprocal of coefficient of variation (mean/stdev) of candidate streaks in L* and S, and std of H; and Textural features are energy, contrast, and homogeneity of candidate streaks. We have also used common color and texture features [4] of the lesion itself (called LCT). LCT includes the following 13 features: The mean, standard deviation and reciprocal of coefficient of variation (mean/stdev) of values in H, S, and V from HSV and L* of L*a*b*, and four of the classical Gray-level co-occurrence matrix based texture measures; energy, contrast, correlation, and homogeneity [4].…”

Section: Feature Extraction and Classificationmentioning

confidence: 99%

“…We have also used common color and texture features [4] of the lesion itself (called LCT). LCT includes the following 13 features: The mean, standard deviation and reciprocal of coefficient of variation (mean/stdev) of values in H, S, and V from HSV and L* of L*a*b*, and four of the classical Gray-level co-occurrence matrix based texture measures; energy, contrast, correlation, and homogeneity [4]. Finally, these 25 features are fed into the SimpleLogistic classifier implemented in Weka.…”

Section: Feature Extraction and Classificationmentioning

confidence: 99%

See 1 more Smart Citation

Oriented Pattern Analysis for Streak Detection in Dermoscopy Images

Medical Image Computing and Computer-Assisted Intervention – MICCAI 2012

Lee

McLean

et al. 2012

Self Cite

Abstract. There is an increasing demand for automated detection and analysis of dermoscopy structures and malignancy clues such as streaks in dermoscopy images, for computer-aided early diagnosis of deadly melanoma. This paper presents a novel approach for streak detection and visualization on dermoscopic images. We tackle the detection of streaks by means of ridge and valley estimation. Orientation estimation and correction is applied to detect low contrast and fuzzy streaks lines, and candidate streaks are used to classify dermoscopy images into streaks Absent or Present with the AUC of 90.5% on 300 dermoscopy images. Our approach can also detect starburst pattern of regular streaks using detected linear structures with accuracy of 81.5% and AUC of 87.7%.