Breast cancer has ranked highest in cancer incidence in Illinois for years. Detection at an early stage helps cancer patients live longer and maintain a better quality of life. Previous research identified two groups of potential risk factors for late-stage breast cancer diagnosis: spatial factors including access to healthcare, and nonspatial factors including socioeconomic and demographic characteristics. Literature suggests that risk factors for late-stage diagnosis behave differently between rural and urban areas, and research needs to separate study areas into various geographic settings. This chapter focuses on an urban area of six counties in Chicago region, and examines possible associations between several risk factors and late-stage breast cancer diagnosis. Based on the data at the zip code level, the study uses the modified scale-space clustering (MSSC) method to form various geographic areas. The MSSC considers both attribute similarity and spatial adjacency while minimizing the loss of information in the clustering process. Therefore, area units defined by the method are more coherent in terms of attribute and spatial closeness for research than geopolitical units. For instance, health literature often suggests the need to separate a study area into urban, suburban and rural areas or even finer-grained area classifications. The method can be used to generate more meaningful geographic divisions than traditional schemes. The analysis results are generally consistent across multiple area units, demonstrating the effectiveness of the method in mitigating the modifiable areal unit problem (MAUP).