DOI
10.5703/1288284318528
Description
This study investigates potential spatial variations in breast cancer incidence across Indiana counties using geographic information systems (GIS) and machine learning (ML) techniques. Breast cancer continues to be a major public health challenge influenced by unequal access to healthcare, socioeconomic disparities, lifestyle factors, and environmental exposures. By integrating GIS and ML, this research identifies key factors and potential spatial patterns associated with breast cancer incidence across Indiana. The statistical analysis produced a highly significant result (p < 0.001), indicating that the independent variables collectively contribute to explaining breast cancer incidence, though the model’s explanatory power was moderate (adjusted R² = 0.25). Among all variables, distance to the nearest hospital and tobacco expenditure showed significant associations with breast cancer incidence, while other factors such as health insurance coverage were not statistically meaningful. The Koenker (BP) statistic was 5.141, indicating spatial stability and confirming that no distinct spatial variation was observed across counties. Findings show that counties farther from healthcare facilities did not necessarily have higher breast cancer incidence, and higher incidence in urban areas, such as Indianapolis, Lake, and Porter Counties, likely reflect better access to screening and diagnosis rather than higher disease occurrence. These results emphasize the importance of incorporating additional variables, including genetic predispositions, racial and ethnic disparities, and environmental conditions, in future studies. Expanding the analytical framework could enhance understanding of spatial patterns and guide efforts to improve healthcare access and cancer prevention in Indiana.
Assessing Disparities in Breast Cancer Incidence Across Indiana Counties with Geographic Information System and Machine Learning
This study investigates potential spatial variations in breast cancer incidence across Indiana counties using geographic information systems (GIS) and machine learning (ML) techniques. Breast cancer continues to be a major public health challenge influenced by unequal access to healthcare, socioeconomic disparities, lifestyle factors, and environmental exposures. By integrating GIS and ML, this research identifies key factors and potential spatial patterns associated with breast cancer incidence across Indiana. The statistical analysis produced a highly significant result (p < 0.001), indicating that the independent variables collectively contribute to explaining breast cancer incidence, though the model’s explanatory power was moderate (adjusted R² = 0.25). Among all variables, distance to the nearest hospital and tobacco expenditure showed significant associations with breast cancer incidence, while other factors such as health insurance coverage were not statistically meaningful. The Koenker (BP) statistic was 5.141, indicating spatial stability and confirming that no distinct spatial variation was observed across counties. Findings show that counties farther from healthcare facilities did not necessarily have higher breast cancer incidence, and higher incidence in urban areas, such as Indianapolis, Lake, and Porter Counties, likely reflect better access to screening and diagnosis rather than higher disease occurrence. These results emphasize the importance of incorporating additional variables, including genetic predispositions, racial and ethnic disparities, and environmental conditions, in future studies. Expanding the analytical framework could enhance understanding of spatial patterns and guide efforts to improve healthcare access and cancer prevention in Indiana.