Pattern Classification of Images from Acetic Acid–Based Cervical Cancer Screening: A Review
Vidya Kudva, Keerthana Prasad · Critical Reviews in Biomedical Engineering · 2018
Automated analysis of digital cervix images acquired during visual inspection with acetic acid (VIA) is found to be of great help to physicians in diagnosing cervical cancer. Application of 3-5% acetic acid to the cervix turns abnormal lesions white, while normal lesions remain unchanged. Digital images of the cervix can be acquired during VIA procedure and can be analyzed using image-processing algorithms. Three main attributes to be considered for analysis are color, vascular patterns, and lesion margins, which differentiate between normal and abnormal lesions. This paper provides a review of state-of-the-art image analysis methods to process digital images of the cervix, acquired during VIA procedure for cervical cancer screening of classification of abnormal lesions.