Open AccessSkin Cancer Diagnosis using SP-SIFT and Machine Learning
Author(s) -
Miss. Divya Rajendra Dhamdhere
Publication year - 2021
Publication title -
international journal for research in applied science and engineering technology
Language(s) - English
Resource type - Journals
ISSN - 2321-9653
DOI - 10.22214/ijraset.2021.35285
Subject(s) - computer science , the internet , cancer , field (mathematics) , cad , skin cancer , artificial intelligence , machine learning , world wide web , medicine , engineering drawing , engineering , mathematics , pure mathematics
Today information handling is become an integral part of our day-to-day life as it is available in abundance. The information from the internet can be used for various purposes from Security to Healthcare management. It can decrease the load on a human being to handle the data manually. So, to highlight the importance of information handling to the society we choose medical field where images are in abundance on the internet. The analysis of images for decision making in medical perspective can be a boon to the medical field and decrease the work of the doctor and increase his productivity to a new level. The cancer has become a menace these days and has become a major health problem. Computer-aided diagnosis (CAD) systems can be used to provide an insight to the cancer specialist and help him to diagnose the various stages of cancer using images and CAD. So, we thought of designing a CAD system which will help in cancer detection and handling various stages of cancers using cancer images as the backend of our application. We propose to implement it on skin cancer which is an ever-increasing cancer due to pollution and environment changes. To design a successful CAD system for skin cancer, it has to be a combination of image processing and machine learning technologies together. In our proposed system we are going to accumulate a lot of images from various cancer image databases which are available free on the internet and design a cancer database with images put under various stages of skin cancer. After database we will apply SP-SIFT algorithm on it and extract features. This database has to available remotely so we have to maintain this feature set on the cloud. The extracted features will be used to train and get results from SVM and Naïve Bayes. Thus, our results will be divided in three categories melanoma, non-melanoma and benign which define stages of skin cancer. Thus, our system will help and improve the decision making and productivity of a medical practitioner.