AI-assisted abnormal CXR findings and correlation with behavioral risk factors: A Public Health Radiography approach to formulating policies and effective interventions

Abstract

Introduction: Cardiovascular, respiratory and related diseases (CVRDs) constitute over 40% cause of death worldwide, mostly reported in low-and-middle-income countries. The catastrophic effect of this spans across poor health outcomes, severe economic loss and significant societal consequences. Responding to this situation necessitates collective strategy to prevent further deterioration as these conditions are closely related, share common risk factors as well as control measures at the clinical, population and policy levels. Thus, this study is aimed at understanding the distribution of AI-assisted abnormal adult chest X-ray (CXR) and examine relationship with behavioral factors; to lay foundation for planned interventions. Methods: Prospective mixed-methods research, cross-sectional in nature, conducted across six top-rated hospitals in Nigeria, representing the six geopolitical zones of the country via purposive sampling technique. Quantitative aspect involved data collection on demographics and abnormal findings from AI-assisted technology, while Qualitative aspect explored individual’s behavioral choices in relation to risk factors. Informed consent and ethical approval were obtained; SPSS software utilized for descriptive and correlation analysis. Results: Cardiomegaly(15.35%), pleural effusion(14.03%), fibrous opacities(10.43%), pleural capping(8.51%), pulmonary mass(7.91%), apical opacities(7.55%), consolidation(6.59%), infiltration(5.88%) among the sixteen abnormal findings in decreasing order of magnitude. An early onset of these anomalies at 30 years was noted, hitting peak values at 40-44 years. A significant percentage of the population engages in unhealthy lifestyle, found to positively correlate with these anomalies in varying degrees; low education levels, health education gaps, poor income and environmental challenges clearly seen. Conclusion: A Public Health Radiography approach- AI assisted, engaging with empirical evidence provides a novel and valuable strategy in designing effective interventions and policy making to address CVRDs burden.

Ähnliche Arbeiten