Lung Cancer Detection and Analysis Using Data Mining Techniques, Principal Component Analysis and Artificial Neural Network

Kassimu Juma, Ma He, Yue Zhao

Abstract


The successful diagnosis of lung cancer disease in early time increases the percentage of patient survival. Effective ways for predict and treat lung cancer remain challenges due to lack of effective ways of detection the lung nodules which causes by their arbitrariness in shape, size and texture. In this paper, image processing is used for image pre-processing, image segmentation and feature extraction. Artificial neural network (ANN) have been employed to learn extracted feature for nodule detection such as shape, size, volume.While principal component analysis were employed for multivariate data processing, it used to detect the complexity of interrelationships between diverse patient, disease and treatment variables.MATLAB have been used for all procedure in processing lung image and artificial neural network for train features extracted. XLSTART software was used for principal component analysis. The lung cancer database which contains the images classify lung image into two kinds:1)Normal with no nodule and 2)nodule image such as benign or malignant.Therefore,by using the proposed method the accuracy obtained was 76%.


Keywords


Artificial Neural Network (ANN); Feature Extraction; Lung Database; Principal Component Analysis (PCA); Region of Interest (ROI); Thresholding.

Full Text:

PDF

Refbacks

  • There are currently no refbacks.


 
 
  
 

 

  


About ASRJETS | Privacy PolicyTerms & Conditions | Contact Us | DisclaimerFAQs 

ASRJETS is published by (GSSRR).