AN ENSEMBLE TEMPLATE MATCHING AND CONTENT-BASED IMAGE RETRIEVAL SCHEME TOWARDS EARLY STAGE DETECTION OF MELANOMA

Malignant melanoma represents the most dangerous type of skin cancer. In this study we present an ensemble classification scheme, employing the mutual information, the cross-correlation and the clustering based on proximity of image features methods, for early stage assessment of melanomas on plain photography images. The proposed scheme performs two main operations. First, it retrieves the most similar, to the unknown case, image samples from an available image database with verified benign moles and malignant melanoma cases. Second, it provides an automated estimation regarding the nature of the unknown image sample based on the majority of the most similar images retrieved from the available database. Clinical material comprised 75 melanoma and 75 benign plain photography images collected from publicly available dermatological atlases. Results showed that the ensemble scheme outperformed all other methods tested in terms of accuracy with 94.9±1.5%, following an external cross-validation evaluation methodology. The proposed scheme may benefit patients by providing a second opinion consultation during the self-skin examination process and the physician by providing a second opinion estimation regarding the nature of suspicious moles that may assist towards decision making especially for ambiguous cases, safeguarding, in this way from potential diagnostic misinterpretations.

[1]  Nikolas P. Galatsanos,et al.  A similarity learning approach to content-based image retrieval: application to digital mammography , 2004, IEEE Transactions on Medical Imaging.

[2]  Wanying Xu,et al.  An efficient template matching between rotated mono- or multi-sensor images , 2011, International Symposium on Multispectral Image Processing and Pattern Recognition.

[3]  C. Ingvar,et al.  Long term results of a randomized study by the Swedish Melanoma Study Group on 2‐cm versus 5‐cm resection margins for patients with cutaneous melanoma with a tumor thickness of 0.8–2.0 mm , 2000, Cancer.

[4]  Constantinos G. Loukas,et al.  Breast Cancer Characterization Based on Image Classification of Tissue Sections Visualized under Low Magnification , 2013, Comput. Math. Methods Medicine.

[5]  T Lee,et al.  Dullrazor®: A software approach to hair removal from images , 1997, Comput. Biol. Medicine.

[6]  N. Cascinelli,et al.  Narrow excision (1-cm margin). A safe procedure for thin cutaneous melanoma. , 1991, Archives of surgery.

[7]  Magnus Snorrason,et al.  Validation of a Skin-Lesion Image-Matching Algorithm Based on Computer Vision Technology. , 2016, Telemedicine journal and e-health : the official journal of the American Telemedicine Association.

[8]  Carlo Tomasi,et al.  Image Similarity Using Mutual Information of Regions , 2004, ECCV.

[9]  A. Tenenhaus,et al.  Detection of melanoma from dermoscopic images of naevi acquired under uncontrolled conditions , 2010, Skin research and technology : official journal of International Society for Bioengineering and the Skin (ISBS) [and] International Society for Digital Imaging of Skin (ISDIS) [and] International Society for Skin Imaging.

[10]  Robert B. Fisher,et al.  A Color and Texture Based Hierarchical K-NN Approach to the Classification of Non-melanoma Skin Lesions , 2013 .

[11]  Raymond L. Barnhill,et al.  Long-Term Results of a Prospective Surgical Trial Comparing 2 cm vs. 4 cm Excision Margins for 740 Patients With 1–4 mm Melanomas , 2001, Annals of Surgical Oncology.

[12]  M E Ming,et al.  The histopathologic misdiagnosis of melanoma: sources and consequences of "false positives" and "false negatives". , 2000, Journal of the American Academy of Dermatology.

[13]  Ophir Schein,et al.  [Effect of dermoscopy on diagnostic accuracy of pigmented skin lesions emphasizing malignant melanoma]. , 2009, Harefuah.

[14]  Ezzeddine Zagrouba,et al.  A PRELIMARY APPROACH FOR THE AUTOMATED RECOGNITION OF MALIGNANT MELANOMA , 2011 .

[15]  J M Elwood.  Skin self-examination and melanoma. , 1996, Journal of the National Cancer Institute.

[16]  Larry D. Hostetler,et al.  The estimation of the gradient of a density function, with applications in pattern recognition , 1975, IEEE Trans. Inf. Theory.

[17]  Jiri Matas,et al.  On Combining Classifiers , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[18]  D. Wan,et al.  Effect of misdiagnosis on the prognosis of anorectal malignant melanoma , 2010, Journal of Cancer Research and Clinical Oncology.

[19]  Field Lm,et al.  Clinical misdiagnosis of melanoma as well as squamous cell carcinoma masquerading as seborrheic keratosis. , 1994 .

[20]  David Polsky,et al.  Early diagnosis of cutaneous melanoma: revisiting the ABCD criteria. , 2004, JAMA.

[21]  Eiliv Lund,et al.  Response: errors in assessing risk factors for melanoma. , 2009, Melanoma research.

[22]  Chun-hung Li,et al.  Minimum cross entropy thresholding , 1993, Pattern Recognit..

[23]  Antonio Soriano Payá,et al.  A decision support system for the diagnosis of melanoma: A comparative approach , 2011, Expert Syst. Appl..

[24]  A. McMichael,et al.  Solar Ultraviolet Radiation: Global burden of disease from solar ultraviolet radiation , 2006 .

[25]  U. Ringborg,et al.  Resection margins of 2 versus 5 cm for cutaneous malignant melanoma with a tumor thickness of 0.8 to 2.0 mm: A randomized study by the Swedish Melanoma Study Group , 1996, Cancer.

[26]  Olaf Gefeller,et al.  Errors in assessing risk factors for melanoma: lack of reproducibility is the minor problem. , 2008, Melanoma research.

[27]  J. Wolf,et al.  Diagnostic inaccuracy of smartphone applications for melanoma detection. , 2013, JAMA dermatology.

[28]  G. Shahgholian,et al.  Robust Object Tracking Using Regional Mutual Information and Normalized Cross Correlation , 2012, 2012 Fourth International Conference on Computational Intelligence and Communication Networks.

[29]  C. Grin,et al.  The misdiagnosis of malignant melanoma. , 1999, Journal of the American Academy of Dermatology.

[30]  Alan Geller,et al.  Methods of Melanoma Detection. , 2016, Cancer treatment and research.

[31]  Joseph Y. Lo,et al.  Mutual information-based template matching scheme for detection of breast masses: From mammography to digital breast tomosynthesis , 2011, J. Biomed. Informatics.

[32]  Anil K. Jain,et al.  Algorithms for Clustering Data , 1988 .

[33]  H F Lorentzen,et al.  Structural asymmetry as a dermatoscopic indicator of malignant melanoma – a latent class analysis of sensitivity and classification errors , 2001, Melanoma research.

[34]  Jacob Scharcanski,et al.  A two-stage approach for discriminating melanocytic skin lesions using standard cameras , 2013, Expert Syst. Appl..

[35]  Vijander Singh,et al.  An efficient similarity measure technique for medical image registration , 2012 .

[36]  P. Carli,et al.  Melanoma detection rate and concordance between self‐skin examination and clinical evaluation in patients attending a pigmented lesion clinic in Italy , 2002, The British journal of dermatology.

[37]  Xiang Li,et al.  A Query-by-Example Content-Based Image Retrieval System of Non-melanoma Skin Lesions , 2009, MCBR-CDS.

[38]  John Paoli,et al.  Skin self-examination using smartphone photography to improve the early diagnosis of melanoma. , 2015 .

[39]  Geoffrey J McLachlan,et al.  Selection bias in gene extraction on the basis of microarray gene-expression data , 2002, Proceedings of the National Academy of Sciences of the United States of America.

[40]  M Stringa.  Misdiagnosis of choroidal melanoma. , 1988, Panminerva medica.

[41]  D. Roberts,et al.  Caution in melanoma risk analysis with smartphone application technology , 2012, The British journal of dermatology.

[42]  J. Abdul Jaleel,et al.  Artificial Neural Network Based Detection of Skin Cancer , 2012 .

[43]  Carlo Riccardo Rossi,et al.  Melanoma: epidemiology, risk factors, pathogenesis, diagnosis and classification. , 2014, In vivo.

[44]  Allan Halpern,et al.  Diagnostic inaccuracy of smartphone applications for melanoma detection: representative lesion sets and the role for adjunctive technologies. , 2013, JAMA dermatology.

[45]  Ioannis Kalatzis,et al.  Computer-based image analysis system designed to differentiate between low-grade and high-grade laryngeal cancer cases. , 2013, Analytical and quantitative cytopathology and histopathology.

[46]  J. Emery,et al.  Smartphone applications for melanoma detection by community, patient and generalist clinician users: a review , 2015, British Journal of Dermatology.

[47]  John W Kelly,et al.  Liability in the context of misdiagnosis of melanoma in Australia , 2014, The Medical journal of Australia.

[48]  Ilias Maglogiannis,et al.  A medical ontology for intelligent web-based skin lesions image retrieval , 2011, Health Informatics J..

[49]  L M Field,et al.  Clinical misdiagnosis of melanoma as well as squamous cell carcinoma masquerading as seborrheic keratosis. , 1994, The Journal of dermatologic surgery and oncology.

[50]  Rafael C. González,et al.  Local Determination of a Moving Contrast Edge , 1985, IEEE Transactions on Pattern Analysis and Machine Intelligence.