Publication Date: 2023/10/07
Abstract: The illness that claims the most lives is lung cancer. It begins with the tissues that are responsible for breathing. The majority of cancer patients have lung cancer, and the survival rate is the same for men and women. Lung cancer is most commonly brought on by smoking, however there are several industrial asbestos products that can harm our lungs and result in lung cancer. Lung cancer can fall into one of two types, the first of which is benign, which is thought to be caused by malignant cells but is less hazardous since it can be treated. It is the earliest stage of lung cancer and only affects a small number of tissues. The second type is malignant, which is serious and deadly and can kill a person. It is strongly advised to begin receiving therapies as soon as possible becauseit is the second and final stage of lung cancer, which is scarcely treatable. Numerous researchers have studied it and attempted to develop a way to identify it in its early stages, while cells are still in the benign state. Lung cancer may be accurately identified through image processing, which is a field of study. There are several methods for implementing it, but accuracy and false alarm rate are important. System should take false alarm rates very seriously and should have high, reliable accuracy. The shape of malignant cells and blood vessels are extracted using Prewitt edge detection in the suggested approach. The proposed approach employs Support Vector Machine (SVM) as well to identify normal and pathological cells so that it can decide right away which grade each one belongs to. The IQ-OTHNCCD dataset, which has a total of 1097 pictures for benign, malignant, and normal categories, was used to evaluate the proposed approach. Compared to previously implemented systems, the system maintained a high degree of accuracy.
Keywords: Lung Cancer, Prewitt Edge Detection, SupportVector Machine, Benign, Malignant, CT-Scan.
DOI: https://doi.org/10.5281/zenodo.8416845
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT23JUL886.pdf
REFERENCES