ABSTRACT
The urgent need for the creation of cutting-edge diagnostic technologies for early detection and improved patient outcomes arises from the fact that lung cancer continues to be the most prevalent and deadly cancer in the world. The techniques of Machine Learning (ML) and Deep Learning (DL) used in the analysis and identification of medical images have shown incredible potential in recent years. This chapter endeavors to explore the potential of these cutting-edge technologies to improve lung cancer identification and classification. This study’s primary goal is to conduct a comparison of ML and DL algorithms in the context of lung cancer detection. A diverse dataset comprising radiological images, histopathological samples, and clinical data is utilized and explored. The various feature selection and fusion processes are carried out to optimize the performance of the models and achieve higher accuracy. Various DL methods and transfer learning techniques are investigated to leverage pre-trained models and bridge the data gap. To overcome potential challenges such as data imbalance, a thorough investigation of techniques to address this issue is conducted, including data augmentation and resampling methods. The study also investigates how genetic biomarkers might be used with imaging data to enhance the precision of lung cancer subtype prediction. The chapter advances the understanding of medical image processing and the detection of lung cancer.
