Please login first
An Analysis of Machine Learning and Image Processing Techniques for Early Detection of Lung Cancer
* , *
1  Dr. Babasaheb Ambedkar Technological University, Lonere, Dist.: Raigad. Maharashtra INDIA.
Academic Editor: Nunzio Cennamo (registering DOI)

Lung cancer is a significant global health concern, necessitating accurate and reliable methods for its diagnosis and classification. This survey paper aims to provide a comprehensive overview of the existing research on lung cancer, focusing on the advancements in diagnostic techniques and classification models. Through a systematic literature review, various machine learning algorithms employed for lung cancer classification were examined, highlighting their strengths and limitations. Additionally, the impact of handling Dicom images on accuracy levels was investigated, emphasizing the need for proper image processing techniques.

The survey reveals that while several classifiers have demonstrated promising results, achieving close to 100% accuracy remains a challenge. Furthermore, the study emphasizes the effectiveness of ensemble classifiers in outperforming other algorithms. To enhance accuracy levels and gain meaningful insights for tumor diagnosis, the paper suggests the development and application of more sophisticated models. Lastly, it emphasizes the significance of further research in the field of Oncology to enhance the classification of benign and malignant lung tumors. This survey paper serves as a valuable resource for researchers, clinicians, and practitioners working towards improved lung cancer diagnosis and classification.

Lung cancer is a formidable global health challenge, accounting for a substantial number of cancer-related deaths worldwide. It is characterized by uncontrolled growth of abnormal cells in the lung tissues, leading to the formation of tumors that can interfere with normal lung function. Lung cancer is a complex and multifaceted disease, with diverse etiological factors and distinct histological subtypes that necessitate comprehensive understanding for effective management and treatment.

The two main types of lung cancer are non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC). NSCLC comprises approximately 85% of all lung cancer cases and is further classified into three subtypes: adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. Each subtype has unique characteristics and may exhibit different responses to treatments. Adenocarcinoma, the most common subtype, typically arises in the outer regions of the lung and is often associated with genetic mutations such as EGFR and ALK. Squamous cell carcinoma arises in the lining of the bronchial tubes and is frequently linked to smoking. Large cell carcinoma is a less common subtype that lacks specific features observed in the other subtypes.

SCLC, on the other hand, accounts for about 15% of lung cancer cases and is characterized by its rapid growth, early metastasis, and association with smoking. SCLC cells are small in size, and the cancer tends to spread quickly to other organs. Due to its aggressive nature, SCLC often requires a different treatment approach compared to NSCLC.

Understanding the distinct sub-types of lung cancer is crucial as it guides treatment decisions, including surgery, radiation therapy, chemotherapy, targeted therapies, and immunotherapy. Moreover, advancements in molecular profiling and precision medicine approaches have provided new opportunities for personalized treatment strategies based on the specific genetic alterations exhibited by individual lung cancer patients.

Keywords: Lung cancer, machine learning, Small cell lung cancer, Non small cell lung cancer, Adenocarcinoma,