Detection Of Lung Cancer From Ct Image Using Svm Classification And

Compare The Survival Rate Of Patients Using 3d Convolutional Neural

Network(3d Cnn)On Lung Nodules Data Set

ABSTRACT :
Lung cancer is the most important cause of cancer death for both men and women. Early
detection is very important to enhance a patient’s chance for survival of lung cancer. This paper
provides a Computer Aided Diagnosis System (CAD) for early detection of lung cancer nodules
from the chest Computer Tomography (CT) images. They are image pre-processing, extraction
of lung region from chest computer tomography images, segmentation of lung region, feature
extraction from the segmented region, classification of lung cancer as benign or malignant.
Initially total variation based denoising is used for image denoising, and then segmentation is
performed using optimal thresholding and morphological operations. Textural features extracted.
For classification, SVM classifier is used. The main aim of the method is to develop a CAD
(Computer Aided Diagnosis) system for finding the lung cancer using the lung CT images and
classify.

EXISTING SYSTEM:
A set of 500 images are utilised in which 75% data is used for training and 25% to achieve
classification. Here the objective of the study is to enhance the classification accuracy by
applying a hybrid classification Algorithm. Lung cancer can be detected by image processing.

Disadvantages :

• It sometimes finds abnormalities that turn out not to be cancer.

• Cost effectiveness

PROPOSED SYSTEM :
In this we use SVM and KMeans algorithm and then comparing prediction accuracy between them. To
implement this we are use lung cancer images dataset saved inside ‘Dataset’ folder. In this dataset we
have two types of images such as normal and abnormal and then SVM and KMEANS will get train on
above dataset and when we upload new image then SVM will predict whether new image is normal or
abnormal. The dataset is loaded and now click on ‘Read & Split Dataset to Train & Test’ button to split
dataset into train and test parts and application split 80% dataset for training and 20% dataset to test
trained model. The SVM accuracy is 60% and now click on “Execute K-Means Algorithm” button to run
KMEANS algorithm on loaded dataset. By this we know svm has better than k means in prediction.

Advantages :

• Lower chance of dying from lung cancer

• Can speed advent of screening.

5. SYSTEM REQUIREMENTS

5.1 HARDWARE REQUIREMENTS:

• System : Pentium Dual Core.

• Hard Disk : 120 GB.
• Monitor : 15’’ LED
• Input Devices : Keyboard, Mouse
• Ram : 1 GB

5.2 SOFTWARE REQUIREMENTS:

• Operating system : Windows 10

• Coding Language : python

• Tool : PyCharm

• Database : MYSQL

• Server : Flask