The Early Detection of Lung Cancer among Indian Patients using Machine Learning Algorithms

Abstract

Lung cancer causes 1 in 5 cancer related deaths globally. Statistics show that less than 16% of lung cancer cases are detected at an initial stage, leading to a significant number of lung cancer deaths within a year of diagnosis. This emphasises a requirement of improved early diagnostic methodologies for lung cancer. For this primary research, we collected comprehensive data from a prominent Indian hospital for over 2500 Indian patients. Our objective was to employ machine learning(ML) models to identify individuals at a heightened risk of developing lung cancer and more specifically, to identify key symptoms that could guide early diagnosis within the Indian populace. Furthermore, we aimed to assess and compare the diagnostic effectiveness of various machine learning (ML) models using established metrics. By evaluating multiple ML models, our study sought to identify the most optimal approach for early lung cancer detection in the Indian population. Notably, among the evaluated models, the Random Forest model emerged as the most promising. It demonstrated remarkable performance across multiple metrics, ultimately achieving the highest diagnostic accuracy of 93.3%. The significant findings of this study provide a strong foundation for future advancements in lung cancer diagnosis in India.