Illumination Correction for Content Analysis in Uterine Cervix Images

Illumination field inhomogeneity strongly affects the visual appearance of an image. It has a major influence on automatic information extraction within an image and its correction is critical for comparison or model learning across images. In this work a unique medical repository of cervicographic images ("cervigrams") collected by the National Center Institute (NCI), National Institute of Health (NIH) is being addressed. The large diversity of cervix shapes within this database, as well as the acquisition set-up, lead to varying illumination conditions among and within the cervigrams, which hamper their automatic analysis. Illumination correction is therefore one of the first preprocessing steps required prior to the image analysis task. This paper presents a method for illumination correction in cervigrams based on a generalized expectation maximization (GEM) algorithm that interleaves pixels classification with estimation of class distribution and illumination field parameters. For cross-image analysis a normalization of the image dynamic range is conducted, using prior knowledge on cervix tissue intensity distribution. Experimental results are provided and evaluated on a set of 110 cervigrams that were manually labeled by an NCI expert. Unsupervised segmentation as well as initial supervised tissue classification results are presented.

[1]  Jose Jeronimo,et al.  A probabilistic approach to segmentation and classification of neoplasia in uterine cervix images using color and geometric features , 2005, SPIE Medical Imaging.

[2]  Hayit Greenspan,et al.  Automatic detection of specular reflections in uterine cervix images , 2006, SPIE Medical Imaging.

[3]  Sunanda Mitra,et al.  An automated, segmentation-based, rigid registration system for Cervigram/spl trade/ images utilizing simple clustering and active contour techniques , 2004, Proceedings. 17th IEEE Symposium on Computer-Based Medical Systems.

[4]  B. Pogue,et al.  Image analysis for discrimination of cervical neoplasia. , 2000, Journal of biomedical optics.

[5]  Shiri Gordon,et al.  Content analysis of uterine cervix images: initial steps toward content based indexing and retrieval of cervigrams , 2006, SPIE Medical Imaging.

[6]  Solomon Kullback,et al.  Information Theory and Statistics , 1970, The Mathematical Gazette.

[7]  P. Cristoforoni,et al.  Computerized Colposcopy: Results of a Pilot Study and Analysis of Its Clinical Relevance , 1995, Obstetrics and gynecology.

[8]  Koenraad Van Leemput,et al.  Automated model-based bias field correction of MR images of the brain , 1999, IEEE Transactions on Medical Imaging.

[9]  Shiri Gordon,et al.  Automatic landmark detection in uterine cervix images for indexing in a content-retrieval system , 2006, 3rd IEEE International Symposium on Biomedical Imaging: Nano to Macro, 2006..

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

[11]  Qiang Ji,et al.  Texture analysis for classification of cervix lesions , 2000, IEEE Transactions on Medical Imaging.

[12]  P. Erdös,et al.  Interpolation , 1953, An Introduction to Scientific, Symbolic, and Graphical Computation.

[13]  Viara Van Raad,et al.  Frequency Space Analysis of Cer-vical Images Using Short Time Fourer Transform , 2003 .