radioelectronicsИзвестия высших учебных заведений России. РадиоэлектроникаJournal of the Russian Universities. Radioelectronics1993-89852658-4794Saint Petersburg Electrotechnical Universityradioelectronics-68Research ArticleТЕЛЕВИДЕНИЕ И ОБРАБОТКА ИЗОБРАЖЕНИЙTELEVISION AND IMAGE PROCESSINGАвтоматический метод анализа мультиспектральных кольпоскопических изображений для телевизионной системы диагностики рака шейки маткиAutomatic method of colposcopic multi-spectral images analysis for television systems diagnostics of cervical cancerОбуховаН. А.ObukhovaN. A.natalia172419@yandex.ruМотыкоА. А.MotykoA. A.motyko.alexandr@yandex.ruСанкт-Петербургский государственный электротехнический университет "ЛЭТИ" им. В. И. Ульянова (Ленина)РоссияSaint Petersburg Electrotechnical University "LETI"Russian Federation201528122015062433Copyright © Обухова Н.А., Мотыко А.А., 20152015Обухова Н.А., Мотыко А.А.Obukhova N.A., Motyko A.A.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://re.eltech.ru/jour/article/view/68

Предложен автоматический метод анализа флуоресцентных изображений, полученных при возбуждающих излучениях с длиной волны 360 и 390 нм. Метод позволяет выявить состояния тканей шейки матки: норму, воспалительный процесс (chronic nonspecific inflammation - CNI), и онкологические изменения (cervical intraepithelial neoplasia - CIN), и построить дифференциальную карту патологии. Для границы CIN/CNI достигнуты чувствительность 87 % и специфичность 71 %. Метод включает специальную предобработку исходных изображений: совмещение изображений, полученных в разных условиях освещения и выделение области интереса. Особенностями метода являются использование совокупности признаков, рассчитанных по изображениям разного типа, и решающее правило при классификации на основе методов интеллектуального анализа данных.

Automated method of fluorescence images analysis obtained by excitation radiation with a wavelength of 360 and 390 nm is proposed. The method allows to detect the status of tissues of cervix: normal, chronic nonspecific inflammation (CNI) and cervical intraepithelial neoplasia (CIN), and build differential map pathology. For the border CIN/CNI achieved a sensitivity of 87 % and specificity 71 %. The method includes a specific preprocessing of the original images: combining images taken in different lighting conditions and highlight the area of interest. Features of the method are the use of a combination of features calculated for images of different types, and decision rule for classification based data mining techniques.

Обработка мультиспектральных изображенийсовмещение мультиспектральных изображенийвыделение области интересаклассификацияобработка медицинских изображенийMulti-spectral images processingmulti-spectral images combiningregion of interest selectionclassificationmedical images processingReferencesCancer Facts & Figures 2006 // URL: http://www. can-cer.org/cancer/cervicalcancer/detailedguide/cervical-cancer-key-statistics (дата посещения 10.12.2015).Cancer Facts & Figures 2006 // URL: http://www. can-cer.org/cancer/cervicalcancer/detailedguide/cervical-cancer-key-statistics (дата посещения 10.12.2015).Cervical cancer incidence and survival in Korea: 1993-2002 / H. H. Chung, M. J. Jang, K. W. Jung et al. // Int. J. of Gynecological Cancer. 2006. Vol. 16, № 5, Р. 1833-1838.Cervical cancer incidence and survival in Korea: 1993-2002 / H. H. Chung, M. J. Jang, K. W. Jung et al. // Int. J. of Gynecological Cancer. 2006. Vol. 16, № 5, Р. 1833-1838.Automated image analysis of digital colposcopy for the detection of cervical neoplasia / S. Y. Park, M. Follen, A. Milbourne et al. // J. of Bio-medical Optics. 2008. Vol. 13, iss. 1. P. (014029-1)-( 014029-10).Automated image analysis of digital colposcopy for the detection of cervical neoplasia / S. Y. Park, M. Follen, A. Milbourne et al. // J. of Bio-medical Optics. 2008. Vol. 13, iss. 1. P. (014029-1)-( 014029-10).Colposcopy for the diagnosis of squamous intraepithelial lesions: a meta-analysis / M. F. Mitchell, D. Schottenfeld, G. Tortolero-Luna et al. // Obstet. Gynecol. 1998. Vol. 91. P. 626-631.Colposcopy for the diagnosis of squamous intraepithelial lesions: a meta-analysis / M. F. Mitchell, D. Schottenfeld, G. Tortolero-Luna et al. // Obstet. Gynecol. 1998. Vol. 91. P. 626-631.Srinvasan Y., Hernes D., Tulpue B. A Probabilistic Approach to Segmentation and Classification of Neoplasia in Uterine Cervix Images Using Color and Geometric Features // SPIE Proc. 2009. Vol. 7259: Medical Imaging 2009: Image Processing // http://spie.org/Publications /Proceedings/Paper/10.1117/12.811210# (дата посещения 10.12.2015).Srinvasan Y., Hernes D., Tulpue B. A Probabilistic Approach to Segmentation and Classification of Neoplasia in Uterine Cervix Images Using Color and Geometric Features // SPIE Proc. 2009. Vol. 7259: Medical Imaging 2009: Image Processing // http://spie.org/Publications /Proceedings/Paper/10.1117/12.811210# (дата посещения 10.12.2015).Xiong J. L.б Wang J. Gu. Image Segmentation of the Acetowhite region in Cervix Images Based on Chromaticity // Proc. of 9 Int. Conf. on In-formation Technology an Applications in Biomedicine (ITAB 2009). Larnaca, Cyprus, 5-7 Nov. 2009. Piscataway: IEEE, 2009. P. 140-144.Xiong J. L.б Wang J. Gu. Image Segmentation of the Acetowhite region in Cervix Images Based on Chromaticity // Proc. of 9 Int. Conf. on In-formation Technology an Applications in Biomedicine (ITAB 2009). Larnaca, Cyprus, 5-7 Nov. 2009. Piscataway: IEEE, 2009. P. 140-144.Huang X., Wang W., Xue Z. Tissue Classification using Cluster Features for Lesion Detection in Digital Cervigrams // SPIE Proc. 2008. Vol. 6914. Medical Imaging 2008: Image Processing // http://spie.org/Publica-tions /Proceedings/Paper/10.1117/12.771088 (дата посещения 10.12.2015).Huang X., Wang W., Xue Z. Tissue Classification using Cluster Features for Lesion Detection in Digital Cervigrams // SPIE Proc. 2008. Vol. 6914. Medical Imaging 2008: Image Processing // http://spie.org/Publica-tions /Proceedings/Paper/10.1117/12.771088 (дата посещения 10.12.2015).Alush A., Greenspan H., Goldberger J. Auto-mated and Interactive Lesion Detection an Segmentation in Uterine Cervix Images // IEEE Trans. on Medical Imaging. 2010. Vol. MI-29, № 2. P. 488-501.Alush A., Greenspan H., Goldberger J. Auto-mated and Interactive Lesion Detection an Segmentation in Uterine Cervix Images // IEEE Trans. on Medical Imaging. 2010. Vol. MI-29, № 2. P. 488-501.Alush A., Greenspan H., Goldberger J. Lesion Detection and Segmentation in uterine cervix images using an arc-level MRF // Proc. of Int. Conf. on Information Technology and Applications in Biomedicine (ISBI 2009). Boston, MA, USA, June 28 - July 1, 2009. P. 474-477 // http://www.sciweavers.org/read/lesion-detection-and-seg mentation-in-uterine-cervix-images-using-an-arc-level-mrf-61381 (дата посещения 10.12.2015).Alush A., Greenspan H., Goldberger J. Lesion Detection and Segmentation in uterine cervix images using an arc-level MRF // Proc. of Int. Conf. on Information Technology and Applications in Biomedicine (ISBI 2009). Boston, MA, USA, June 28 - July 1, 2009. P. 474-477 // http://www.sciweavers.org/read/lesion-detection-and-seg mentation-in-uterine-cervix-images-using-an-arc-level-mrf-61381 (дата посещения 10.12.2015).Content Analysis of Uterine Cervix Images: Initial Steps Towards Content Based Indexing and Retrieval of Cervigrams // S. Gordon, G. Zimmer-man, R. Long et al. // SPIE Proc. 2006. Vol. 6144. Medical Imaging 2006: Image Processing. P. 1549-1556.Content Analysis of Uterine Cervix Images: Initial Steps Towards Content Based Indexing and Retrieval of Cervigrams // S. Gordon, G. Zimmer-man, R. Long et al. // SPIE Proc. 2006. Vol. 6144. Medical Imaging 2006: Image Processing. P. 1549-1556.Automated image analysis of digital colposcopy for the detection of cervical neoplasia / S. Park, M. Follen, A. Milbourne et al. // J. Biomed Opt. 2008. Vol. 13, № 1. P. 14-29.Automated image analysis of digital colposcopy for the detection of cervical neoplasia / S. Park, M. Follen, A. Milbourne et al. // J. Biomed Opt. 2008. Vol. 13, № 1. P. 14-29.Multispectral digital colposcopy for in vivo detection of cervical cancer / J. M. Benavides, S. Chang, S. Y. Park et al. // Optics express. 2003. Vol. 11, № 10. P. 1223-1236.Multispectral digital colposcopy for in vivo detection of cervical cancer / J. M. Benavides, S. Chang, S. Y. Park et al. // Optics express. 2003. Vol. 11, № 10. P. 1223-1236.Multispectral digital microscopy for in vivo detection of oral neoplasia in the hamster cheek pouch model of carcinogenesis / S. Y. Park, K. Sokolov, T. Collier et al. // Optics Express. 2005. Vol. 13. P. 749-762.Multispectral digital microscopy for in vivo detection of oral neoplasia in the hamster cheek pouch model of carcinogenesis / S. Y. Park, K. Sokolov, T. Collier et al. // Optics Express. 2005. Vol. 13. P. 749-762.Van Raad V., Bradley A. B. Emerging Technologies, Signal proceeding and statistical methods for screening of cervical cancer in vivo - are they good candidates for cervical screening? // Proc. of IEEE 2nd Int. Conf. on Advances in Medical Signal and Information Processing (MEDSIP'04), Malta, 5-8 Sept., 2004. P. 210-217 // http://espace.lib-rary.uq.edu.au/view/UQ:8932/MEDSIP04 _VVR_APB.pdf (дата посещения 10.12.2015).Van Raad V., Bradley A. B. Emerging Technologies, Signal proceeding and statistical methods for screening of cervical cancer in vivo - are they good candidates for cervical screening? // Proc. of IEEE 2nd Int. Conf. on Advances in Medical Signal and Information Processing (MEDSIP'04), Malta, 5-8 Sept., 2004. P. 210-217 // http://espace.lib-rary.uq.edu.au/view/UQ:8932/MEDSIP04 _VVR_APB.pdf (дата посещения 10.12.2015).Ji Q., Engel J., Craine E. Texture Analysis for classification of Cervix Lesions // IEEE Trans. on Medical Imaging. 2010. Vol. 19, № 11. P. 1144-1149.Ji Q., Engel J., Craine E. Texture Analysis for classification of Cervix Lesions // IEEE Trans. on Medical Imaging. 2010. Vol. 19, № 11. P. 1144-1149.Ji Q., Engel J., Craine E. Classifying cervix tissue patterns with texture analysis // Pattern Recognition. 2000. Vol. 33, № 9. P. 1561-1573.Ji Q., Engel J., Craine E. Classifying cervix tissue patterns with texture analysis // Pattern Recognition. 2000. Vol. 33, № 9. P. 1561-1573.Segmentation and Classification of Cervix Lesions by Pattern and Texture Analysis / Bh. Tul-pule , Sh. Yang, Y. Srinivasan et al. // Proc. of IEEE Int. Conf. on Fuzzy System. 22-25 May 2005. Reno, NV, USA. Piscataway: IEEE, 2005. P. 173-176.Segmentation and Classification of Cervix Lesions by Pattern and Texture Analysis / Bh. Tul-pule , Sh. Yang, Y. Srinivasan et al. // Proc. of IEEE Int. Conf. on Fuzzy System. 22-25 May 2005. Reno, NV, USA. Piscataway: IEEE, 2005. P. 173-176.Tulpule B. Color and Texture Analysis of Cervix Lesions: MS in Electrical Engineering Degree Thesis // Texas Tech. University. Lubbock, TX, USA. 2004, 56 p. // https://www.google.ru/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwjvndncyaLKAhVH_3IKHTqsAQ4QFggfMAA&url=https%3A%2F%2Frepositories.tdl.org%2Fttuir%2Fbitstream%2Fhandle%2F2346%2F13541%2F31295019800795.pdf%3Fsequence%3D1&usg=AFQjCNFYL9wqDUi-_ovmDIWOMfaf0B9znA&cad=rjt (дата посещения 10.12.2015).Tulpule B. Color and Texture Analysis of Cervix Lesions: MS in Electrical Engineering Degree Thesis // Texas Tech. University. Lubbock, TX, USA. 2004, 56 p. // https://www.google.ru/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwjvndncyaLKAhVH_3IKHTqsAQ4QFggfMAA&url=https%3A%2F%2Frepositories.tdl.org%2Fttuir%2Fbitstream%2Fhandle%2F2346%2F13541%2F31295019800795.pdf%3Fsequence%3D1&usg=AFQjCNFYL9wqDUi-_ovmDIWOMfaf0B9znA&cad=rjt (дата посещения 10.12.2015).Li Wenjing, Poirson A. Detection and Characterization of Abnormal Vascular Patterns in Automated Cervical Image Analysis // URL: http://www.sti-hawaii.com/docs /ISVC2006 /ISVC06_225_CameraReady.pdf (дата посещения 10.12.2015).Li Wenjing, Poirson A. Detection and Characterization of Abnormal Vascular Patterns in Automated Cervical Image Analysis // URL: http://www.sti-hawaii.com/docs /ISVC2006 /ISVC06_225_CameraReady.pdf (дата посещения 10.12.2015).Classification of Cervix Lesions Using Filter Bank-Based Texture Models / Y. Srinivasan, B. Nutter, S. Mitra et al. // Proc. of the 19th IEEE Symp. Computer-Based Medical Systems (CBMS’06). 22-23 June 2006, Salt Lake City, UT, USA. Piscataway: IEEE, 2006. P. 832-840.Classification of Cervix Lesions Using Filter Bank-Based Texture Models / Y. Srinivasan, B. Nutter, S. Mitra et al. // Proc. of the 19th IEEE Symp. Computer-Based Medical Systems (CBMS’06). 22-23 June 2006, Salt Lake City, UT, USA. Piscataway: IEEE, 2006. P. 832-840.Gordon S., Zimmerman G., Greenspan H. Image segmentation of Uterine Cervix images for indexing in PACS // Proc. of the 17th IEEE Symp. on Computer-Based Medical Systems (CBMS’04). 24-25 June 2004, Tel Aviv, Israel. Piscataway: IEEE, 2004. P. 298-303.Gordon S., Zimmerman G., Greenspan H. Image segmentation of Uterine Cervix images for indexing in PACS // Proc. of the 17th IEEE Symp. on Computer-Based Medical Systems (CBMS’04). 24-25 June 2004, Tel Aviv, Israel. Piscataway: IEEE, 2004. P. 298-303.Комплекс для флуоресцентной диагностики и фотодинамической терапии заболеваний шейки матки / Н. А. Обухова, Г. В. Папаян, А. А. Мотыко и др. // Опт. журнал. 2015. Т. 82, № 12. С. 47-59.Комплекс для флуоресцентной диагностики и фотодинамической терапии заболеваний шейки матки / Н. А. Обухова, Г. В. Папаян, А. А. Мотыко и др. // Опт. журнал. 2015. Т. 82, № 12. С. 47-59.Stone H. S. A Fast Direct Fourier-Based Algorithm for Subpixel Registration of Images // IEEE Trans. on Geoscience and Remote Sensing. 2001. Vol. GRS-39, № 10. P. 2235-2242.Stone H. S. A Fast Direct Fourier-Based Algorithm for Subpixel Registration of Images // IEEE Trans. on Geoscience and Remote Sensing. 2001. Vol. GRS-39, № 10. P. 2235-2242.Das A., Kar A., Bhattacharyya D. Elimination of Specular reflection and Identification of ROI: The First Step in Automated Detection of Uterine Cervical Cancer using Digital Colposcopy // IEEE Int. Conf. on Imaging Systems & Techniques (IST’11). 17 May 2011, Penang, Malaysia. Piscataway: IEEE, 2011. P. 237-141.Das A., Kar A., Bhattacharyya D. Elimination of Specular reflection and Identification of ROI: The First Step in Automated Detection of Uterine Cervical Cancer using Digital Colposcopy // IEEE Int. Conf. on Imaging Systems & Techniques (IST’11). 17 May 2011, Penang, Malaysia. Piscataway: IEEE, 2011. P. 237-141.Breiman L. Random Forests // J. Machine Learning. 2001. Vol. 45, iss. 1. P. 5-32.Breiman L. Random Forests // J. Machine Learning. 2001. Vol. 45, iss. 1. P. 5-32.Meinshausen N. Quantile Regression Forest // J. Machine Learning Research. 2006. Vol. 7, iss. 6. P. 983-999.Meinshausen N. Quantile Regression Forest // J. Machine Learning Research. 2006. Vol. 7, iss. 6. P. 983-999.Han J., Kamber M. Data mining: concepts and techniques. 2nd. ed. San Francisco: Morgan Kaufmann, Elsevier Science, 2006. 137 pHan J., Kamber M. Data mining: concepts and techniques. 2nd. ed. San Francisco: Morgan Kaufmann, Elsevier Science, 2006. 137 p

The authors declare that there are no conflicts of interest present.