Early Detection of Lung Cancer of CT Scans in Biomedical Image Processing Using Feature Extraction Methods and Support Vector Machine (SVM) Classification

Main Article Content

Retz Mahima Devarapalli ,Sajja Tulasi Krishna ,Hemantha Kumar kalluri

Abstract

Introduction: Lung cancer remains a significant cause of mortality, underscoring the critical need for early detection to improve survival rates. Despite existing methods for lung cancer detection, accuracy enhancements remain a priority. In this paper, we propose a novel approach utilizing Fuzzy-C-Means (FCM) and Support Vector Machine (SVM) classification for lung cancer detection.


Objectives: The primary objective of this study is to develop a robust lung cancer detection method that outperforms existing approaches in terms of accuracy.


Methods: The proposed methodology comprises four key steps:


Pre-processing: Raw images undergo pre-processing using Median Filter (MF) to enhance their quality and reduce noise.


Segmentation: The pre-processed images are segmented using the Fuzzy-C-Means algorithm (FCM), which partitions the image into distinct regions, facilitating subsequent analysis.


Feature Extraction: Local Binary Pattern (LBP) is employed to extract discriminative features from the segmented images. LBP is known for its effectiveness in capturing texture information, making it suitable for our classification task.


Classification: Extracted features are fed into Support Vector Machine (SVM) for classification. SVM is chosen for its ability to handle high-dimensional data and its robustness in classification tasks.


Results: Experimental evaluations were conducted on two standard benchmark datasets: LIDC Dataset and SPIE-AAPM Dataset. The findings demonstrate the superiority of our proposed approach over state-of-the-art methods. Specifically, our method achieves higher accuracy in lung cancer detection, thereby validating its effectiveness in improving early detection rates.


Conclusions: The proposed approach combining Fuzzy-C-Means segmentation and SVM classification presents a promising solution for enhancing lung cancer detection accuracy. By leveraging advanced image processing techniques and machine learning algorithms, our method demonstrates superior performance compared to existing approaches. These findings underscore the potential of our method to contribute significantly to early detection efforts, ultimately leading to improved patient outcomes in the fight against lung cancer.

Article Details

Section
Articles