The role of machine learning in lung cancer prediction: Insights from a multifactorial risk assessment

Main Article Content

Emek Guldogan

Abstract

Aim: Lung cancer is a multifaceted condition that is affected by a range of lifestyle, environmental, and hereditary factors. The prevalence of lung cancer is on the rise in some areas due to elevated rates of smoking and air pollution. This study aims to investigate the factors contributing to the development and progression of lung cancer, with a specific focus on evaluating the predictive significance of various lifestyle, environmental, and genetic variables.


Materials and Methods: The research used a publically accessible dataset from Kaggle, which consisted of 16 characteristics and 3,310 occurrences. The data included demographic, behavioral and health-related characteristics, including gender, smoking, anxiety, exhaustion, and chronic illness. An MLP model was used to evaluate the predictive importance of each variable. The dataset was split into 70% for training and 30% for testing. The relative effect of factors on lung cancer risk was compared using the normalized importance.


Results: The research demonstrated a robust correlation between lung cancer and smoking, coughing, yellow fingers, and chest discomfort. Additionally, fatigue and allergies were important indicators. Nevertheless, there were no notable disparities in lung cancer occurrence based on gender and age. Age was identified as the primary predictor in the MLP model, with shortness of breath, alcohol intake, yellow fingers and smoking following as subsequent predictors.


Conclusion: The research affirms the well-known correlation between smoking and lung cancer, emphasizing the significance of early indicators such as persistent cough and chest discomfort. The lack of notable gender and age disparities implies that behavioral and symptomatic variables may play a more crucial role in determining the risk of developing lung cancer. The results endorse inclusive lung cancer screening initiatives that take into account other variables, such as environmental exposure and genetic predisposition, in addition to conventional risk factors like smoking.

Downloads

Download data is not yet available.

Article Details

How to Cite
Guldogan, E. (2024). The role of machine learning in lung cancer prediction: Insights from a multifactorial risk assessment. Annals of Medical Research, 31(10), 834–839. Retrieved from http://www.annalsmedres.org/index.php/aomr/article/view/4762
Section
Original Articles