Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/534323
Title: | Performance analysis of automated classification of cervical cancer cells using pap smear |
Researcher: | Lavanya Devi N |
Guide(s): | Aravind, K R |
Keywords: | Cervical Cancer Human Papillomavirus Pap Smear |
University: | Anna University |
Completed Date: | 2023 |
Abstract: | Cervical cancer is one of the most dangerous cancerous which ranks newlinefour among the cancers affected worldwide. The major cause of the disease is newlinethe ignorance about this type of cancer. Human Papillomavirus (HPV) is the newlinemajor cause of cancer. The risk factors for cervical cancer are sex at an early newlineage, weak immune system, usage of drugs, consumption of birth control pills newlineetc. The test carried out for the detection of HPV is a pap smear test. At present, newlinethe pap images are analyzed manually for the detection of cervical cancer. newlineThe manual method of detection is erroneous as the abnormal cells are very newlinesimilar to the normal cells in appearance and as a result, it is subject to false newlinepositive and false negative cases. This research work involves automated detection of cervical cancer from pap images. For the automated detection of cervical cancer machine learning algorithms, neural networks and transfer learning algorithms are used. Geometrical Features (19 Nos.) and texture features (4 Nos.) are extracted to newlineclassify the cell into normal and abnormal cells. If the twenty three features are newlineconsidered for the whole process it will consume time. In order to overcome this newlinelimitation, the four most significant features are identified by applying Principal newlineComponent Analysis (PCA). These significant features are given as the input to newlinethe machine learning algorithms and neural network. Machine learning newlinealgorithms like KNN, Fine Gaussian SVM, Ensemble Bagged trees and Linear newlineDiscriminant are used for the classification of the input images into normal and newlineabnormal. The main limitation of machine learning algorithm is that the features newlineare extracted manually. The accuracy of the system will depend on the features newlineextracted. The detection of cervical cancer is also implemented by neural newlinenetworks. The neural network depicts the working principle of the human brain. newlineThe neural network has input layers, hidden layer and output layer. newline newline |
Pagination: | xvi,134p |
URI: | http://hdl.handle.net/10603/534323 |
Appears in Departments: | Faculty of Information and Communication Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 364.03 kB | Adobe PDF | View/Open |
02_prelim pages.pdf | 2.38 MB | Adobe PDF | View/Open | |
03_content.pdf | 61.62 kB | Adobe PDF | View/Open | |
04_abstract.pdf | 51.17 kB | Adobe PDF | View/Open | |
05_chapter 1.pdf | 787.82 kB | Adobe PDF | View/Open | |
06_chapter 2.pdf | 151.62 kB | Adobe PDF | View/Open | |
07_chapter 3.pdf | 650.79 kB | Adobe PDF | View/Open | |
08_chapter 4.pdf | 825.77 kB | Adobe PDF | View/Open | |
09_annexures.pdf | 132.58 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 70.95 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: