Computer Science and Electrical Engineering Session:

Title:            Automated Identification and Analysis of Cervical Cancer
Authors:       Samya Muhuri, Monalisa Bhattacharjee
Abstract:      Recent advances in imaging technology like USG, CT, MRIs, PET scans and other techniques have a huge impact on the diagnosis and treatment of disease. Medical imaging helps earlier detection of various malignant diseases such as breast cancer, lung cancer, cervical cancer etc. Cervical cancer is the second most common cancer among women in developing countries. Cervical benign lesions and invasive cancer exhibit some morphological features that can be detected during a conventional visual examination. If it can be detected earlier and treated properly then the patient can survive. Computational techniques in the processing of histopathological images allow us to assist the oncologist with a Computer Aided Diagnosis system. The overall objective of the work is to implement and evaluate a method for the analysis of cervical images. Epithelium that turns white after application of acetic acid is called acetowhite epithelium. Acetowhite epithelium is one of the major diagnostic features observed in detecting precursor lesions and cancerous region. Here we present a multistep acetowhite region detection system to analyze lesions in uterine cervical images. First the Gabor filter is used to enhance the image. Second, the uterine cervix is segmented in different regions by using of Marker Control Watershed Segmentation algorithm. Third, acetowhite region are extracted through Support Vector Machine classifier. The above mentioned processes are tested on different uterine cervical images randomly retrieved from popular search engine and work has been implemented by MATLAB R2009b package.

Keywords:    Cervical cancer; Acetowhite region; Computer-aided diagnosis (CAD); Feature extraction; Gabor filter; Support vector machine.
Pages:          516-520

Download Full Paper