A novel humanitarian technology for early detection of cervical neoplasia: ROI extraction and SR detection

Cervical cancer is a common form of cancer in developing countries. Through screening programs it can be preventable. During colposcopy, Specular Reflections(SR) comes as a heavily bright spot and generate problems in the image analysis of the cervix. This occurs due to the present of fluids and the cervix being convex in shape and due to the illumination generated from the colposcope. In this paper, SRs are detected and filled by performing top-hat filtering, global thresholding, different morphological operations and a novel filling algorithm. The main concern is to detect AW regions. The non-ROI portion contains unnecessary information that confuses and creates problems in the detection of different tissues present in the cervix. The ROI is extracted by considering hue (H) and value (V) from HSV color space, so that the boundary of the colposcopy image i.e. the cervix is extracted. Finally our Algorithm on the ROI is compared with the expert markings and statistical metrics are computed. This is an ongoing novel Humanitarian Technology in detection of Cervical Cancer.

[1]  Hayit Greenspan,et al.  Lesion detection and segmentation in uterine cervix images using an ARC-LEVEL MRF , 2009, 2009 IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[2]  Avijit Kar,et al.  Early Detection of Cervical Cancer Using Novel Segmentation Algorithms , 2014 .

[3]  Sun Young Park,et al.  Domain-Specific Image Analysis for Cervical Neoplasia Detection Based on Conditional Random Fields , 2011, IEEE Transactions on Medical Imaging.

[4]  Abhishek Das,et al.  Elimination of specular reflection and identification of ROI: The first step in automated detection of Cervical Cancer using Digital Colposcopy , 2011, 2011 IEEE International Conference on Imaging Systems and Techniques.

[5]  Abhishek Das,et al.  Preprocessing for Automating Early Detection of Cervical Cancer , 2011, 2011 15th International Conference on Information Visualisation.

[6]  Sunanda Mitra,et al.  An automated, segmentation-based, rigid registration system for Cervigram/spl trade/ images utilizing simple clustering and active contour techniques , 2004 .

[7]  Hayit Greenspan,et al.  Automated and Interactive Lesion Detection and Segmentation in Uterine Cervix Images , 2010, IEEE Transactions on Medical Imaging.

[8]  Yen-Ping Chu,et al.  Edge Enhancement Nucleus and Cytoplast Contour Detector of Cervical Smear Images , 2008, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[9]  Shiri Gordon,et al.  SEGMENTATION OF NON-CONVEX REGIONS WITHIN UTERINE CERVIX IMAGES , 2007, 2007 4th IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[10]  Shiri Gordon,et al.  Illumination Correction for Content Analysis in Uterine Cervix Images , 2006, 2006 Conference on Computer Vision and Pattern Recognition Workshop (CVPRW'06).

[11]  A. Kar,et al.  Implication of technology on society in Asia: Automated detection of Cervical Cancer , 2012, 2012 IEEE Conference on Technology and Society in Asia (T&SA).