Home
Registration
Super-Resolution
Bacteria Identification
Doppler analysis
MSL segmentation
CGMM
X-ray classification
Cervigrams
Image Clustering
Video Indexing
Object tracking
Gmm&Edge
DTI

 

Content Analysis of
Uterine Cervix Images (Cervigrams)
for Indexing in CBIR systems

 

Overview

This work is motivated by the need for visual information extraction and management in the growing field of medical image archives. In particular the work focuses on a unique medical repository of digital cervicographic images ("Cervigrams") collected by the National Cancer Institute (NCI) in a longitudinal multi-year studies carried out in Guanacaste, Costa Rica and the united states. NCI together with the National Library of Medicine (NLM) is developing a unique Web-based database of the digitized cervix images to study the evolution of lesions related to cervical cancer. The work is part of an on-going effort towards the creation of a content-based image retrieval (CBIR) system for the cervicographic images. Such a system requires specific tools that can analyze the cervigram content and represent it in a way that can be efficiently searched and compared.

We present a multi-step scheme for segmenting and labeling regions of medical and anatomical interest within the cervigram.  The multi-step structure is motivated by the large diversity of the images within the database. The algorithm initially identifies the cervix region within the image and allocates the cervix boundary and the os landmarks. It than performs illumination correction and intensity normalization of the images across the data-set. This step is shown to improve successive  segmentation results.  Tissue segmentation is handled next and the cervix region is segmented into three main tissue types: the columnar epithelium (CE), the squamous epithelium (SE), and the acetowhite (AW), which is visible for a short time following the application of acetic acid. Areas of acetowhitening correlate with higher nuclear density and are of clinical signifficance.

Results for the different steps within this scheme are evaluated on different sets of manually labeled cervigrams, for which the markings of one to twenty experts are available. When an image is marked by more than one expert  different analysis schemes based on the output of the STAPLE algorithm ( Warfield et.al - TMI 2004) are presented.

 

Publications

bulletS. Gordon, PhD Thesis, submitted on July 2009.  Automatic content analysis of uterine cervix images using computerized tools. [pdf]
bulletS. Gordon, S. Lotenberg, R. Long, S. Antani, J. Jeronimo for the NIH-ASCCP Research Group and H. Greenspan. “Evaluation of Uterine Cervix Segmentations using Ground Truth from Multiple Experts”. Accepted for publication in Computerized Medical Imaging and Graphics (CMIG), December 2008.
bulletH. Greenspan, S. Gordon, G. Zimmerman, S. Lotenberg, J. Jeronimo, S. Antani and R. Long. “Automatic Detection of Anatomical Landmarks in Uterine Cervix Images”. Accepted for publication in IEEE Transaction on Medical Imaging (TMI), August 2008
bulletS. Lotenberg , S. Gordon and H. Greenspan. "Shape priors for segmentation of the cervix region within uterine cervix images". Journal of Digital Imaging, accepted for publication 2008.
bulletS. Lotenberg , S. Gordon and H. Greenspan. "Shape priors for segmentation of the cervix region within uterine cervix images". In Proc. of SPIE medical imaging, 2008. [pdf]
bulletS.Gordon and H. Greenspan. "Segmentation of non-convex regions within uterine cervix images".  In Proc. of  IEEE International Symposium on Biomedical Imaging, 2007. [pdf]
bullet

S. Lotenberg, S. Gordon, R. Long, S. Antani, J. Jeronimo and H. Greenspan. "Automatic evaluation of uterine cervix segmentations". In Proc. of SPIE medical imaging, 2007. [pdf]

bullet

H. Dvir, S. Gordon and H. Greenspan. "Illumination correction for content analysis in uterine cervix images". In Proc. of  IEEE WS Mathematical Methods in Biomedical Image Analysis (MMBIA), Computer Vision and Pattern Recognition Workshop (CVPRW06), page 95, 2006. [pdf]

bullet

G. Zimmerman, S. Gordon  and H. Greenspan. "Automatic landmark detection in uterine cervix images for indexing in a content-retrieval system". In Proc. of IEEE International Symposium on Biomedical Imaging, pages 1348-1351, 2006. [pdf]

bullet

S. Gordon, G. Zimmerman, R. Long, S. Antani, J.Jeronimo and H. Greenspan. "Content analysis of uterine cervix images: Initial steps towards content based indexing and retrieval of cervigrams". In  Proc. of SPIE medical imaging, volume 6144, pages 1549-1556, 2006. [pdf]

bullet

G. Zimmerman and H. Greenspan. "Automatic detection of specular reflections in uterine cervix images". In
Proc. of SPIE Medical Imaging, volume 6144, pages 2037- 2045, 2006.
[pdf]

bullet

S. Gordon , G. Zimmerman and H. Greenspan. "Image segmentation of uterine cervix images for indexing in PACS", In Proc. of 17th IEEE Symposium on Computer-Based Medical Systems, CBMS 2004. Bethesda, MD, 2004. [pdf]

bullet

G. Zimmerman, S. Gordon and H. Greenspan. "Content-based indexing and retrieval of uterine cervix images",  In Proc. of 23rd IEEE Convention of Electrical and Electronics Engineers in Israel 2004, pages 181-185, Tel-Aviv, Israel, 2004. [pdf]

horizontal rule

Home Registration Super-Resolution Bacteria Identification Doppler analysis MSL segmentation CGMM X-ray classification Cervigrams Image Clustering Video Indexing Object tracking Gmm&Edge DTI

For problems or questions regarding this web page contact hayit@eng.tau.ac.il.
Last updated: 01/07/08.