Abnormality Detection and Severity Classification of Cells based on Features Extracted From Papanicolaou Smear Images using Machine Learning

摘要

A Papanicolaou Smear (PAP) test is a screening method developed for cervical cancer that involves the microscopic examination of cervical cells carefully extracted and spread out as a smear and stained specially. A Pap test reveals premalignant and malignant changes and the changes that are due to non-carcinogenic conditions like inflammation. The diagnosis of this test are based upon key features of the nucleus and cytoplasm of the affected cell or the cell under observation. This work is aimed at devising a classification algorithm using supervised methods to efficiently classify the affected cells from normal cells and further group the affected cells Logistic Regression.^{[9] All algorithms and models were trained and validated using the Azure Machine Learning Studio.}