Exploration of Social-economic and Environmental Impacts on Lung Cancer Mortality in 94 Countries With Artificial Neural Network and Random Decision Tree Analysis

Background: Since lung cancer is the biggest killer in cancer families, extremely threatening to human health, an understanding of social-economic and environmental impacts on lung cancer mortality (LCM) is imperative to improve patient psychological health and potentially mitigate lung cancer incidences and multimorbidity. Figuring out key indicators in social-economic and environment impacts which are sensitive to LCM on spatial-temporal scales contributes to preclinical control and systemic treatments for lung cancer with standard chemotherapy agents are still relatively ineffective. Methods: Based on lung cancer mortalities in 94 countries within a decade (2006-2016), this research appropriately dissects social-economic, demographic, and environmental independent variable effects using Artificial Neural Network (ANN) and CRT Random Decision Tree algorithms (CRT-CRT-RDF). Results: With the two methods comparison, the similarity is that education and carbon emission were two etiologies. Education in low and middle countries of lung cancer was related with total ecological footprint and total population. Carbon emission in extreme countries was linked to ecological forestland footprint. Spatial-temporal analysis postulated China and the U.S were the two largest countries whereas China and India were the two fastest countries of LCM growth. Both models have a high precision of prediction (96.1% of CRT-CRT-RDF and 98.4% of ANN). Conclusions: This research will facilitate preventive lung cancer services, prioritize the geographical allocation of lung cancer investment for WHO, and provide evidence for shrinking carbon emission and deforestation.