Revolutionizing Lung Cancer Diagnosis: A Comprehensive Review of Image Processing Techniques for Early Detection and Precision Medicine

Abstract

According to World Health Organisation (WHO), lung cancer is the leading cause of cancer-related fatalities in both genders and has the highest fatality rate. Early detection of pulmonary nodules is essential to improving the significant survival rate of lung cancer due to the typical proliferation of lung cells. Studies on lung cancer indicate that smoking is the primary cause of this disease, which is more common in women nowadays and causes more deaths than breast cancer. Age, gender, race, socioeconomic status, exposure to the environment, air pollution, alcohol consumption, and second-hand smoking are a few more factors that could be significant in causing lung cancer. Early detection of lung cancer is achieved through a variety of image processing techniques, such as computed tomography (CT), bone scanning, magnetic resonance imaging (MRI), Positron Emission Tomography, PET-CT, and X-ray scanning. These techniques are combined with machine learning algorithms, data mining, and artificial intelligence-based detection techniques, which improve detection through efficient computing systems known as computer assisted diagnosis (CAD). Since practically all lung cancer screening and detection is dependent on image processing, this article will serve as a reference for aspiring researchers to understand the many detection strategies in effectively identifying lung cancer. Additionally, five distinct methods are evaluated and critically analysed, along with their benefits and drawbacks, taking into account the present and potential future developments in early lung cancer diagnosis for human survival.