Malignant Cancer Cell Detection By Microscopy Image Analysis
Hussain, Md. Najmul
Rahman, Mohammad Azizur
MetadataShow full item record
This is a kind of image processing work, which help to check the malignancy of a cell by testing a biopsy image and provide faster salutation than traditional diagnosis system. The aim of the system is to check malignancy of a cells biopsy image for this firstly we do color based segmentation. After Color based segmentation we have done Clustering by using k-means algorithm. When clustering is done we have selected the best cluster among them and do feature extraction like nucleus counting and number of white and black pixel. After that we compared it weather it is in the ratio of class normal, invasive or insitu which indicate its malignancy level. To develop this project the most essential tool was MATLAB which helps to simulate all the parts of the work and test it. Another tool that is used in the development work is Java technology. Some other tools as photoshop, Jaspersoft for iReport are used in some portions of the development. After implementation of all functions, the system is tested in different stages and it works successfully as a prototype.