Inflammation Detection in Cervical Cytology Images
Published: 2017
Author(s) Name: K. V. Sheena, K. V. Sikha |
Author(s) Affiliation: Asst. Prof., Computer Science , AdiShankara Institute of Engg. and Tech., Kerala, India.
Locked
Subscribed
Available for All
Abstract
Cervical cancer is a leading cause of cancer-related deaths in women worldwide. If detected at a precancerous condition it is completely curable. Several screening methods are available and Pap smear test is a preliminary screening method. Hence an automatic detection method for cervical cancer will be preferred. Cervical Screening System is a computer assisted screening solution, where the digitized images of the PAP-Smear are analyzed and classified through advanced image processing and classification algorithms. The LBC slides are analyzed for the detection of normal and abnormal cells. But some slides will be inflammatory, that are neither normal nor abnormal, but existing systems categorize them as abnormal and will go for further review. The inflammatory slides should be identified and categorized as Inflammatory so that they will not go for further review, instead they will go for a repap (preliminary) test after 6-9 months. The identification of Inflammatory slides inturn increases the system accuracy which increases specificity and sensitivity.
Keywords: Cervical Screening System, Feature Extraction, Inflammation, Neutrophils, Segmentation, SVM
View PDF