TEMPORAL ASSOCIATION RULES IN BREAST CANCER

Resumo

This study aimed to present the potential of the process of  Knowledge Discovery in Databases  to better understand the evolution of diagnosis up to surgical intervention in women diagnosed with breast cancer who were beneficiaires of health insurance providers. Thus, association rules were discovered and were post-processed to identify possible cause-effect situations. A database of 50 women who underwent at least one conservative or mutilating surgery from 1994 to 2013 was adopted. The post-processed rules, which considered temporal windows, showed that although surgical treatments are performed in short periods of time, the decision on the type of procedure to be used depends on the time to diagnosis, related to mammographic detection routines. The result prompted the need for adjustment in the period of time between surgical procedures and screening tests. The process of  Knowledge Discovery in Databases was found to be useful to support procedures that allow early diagnosis and treatment, aimed to favor the use of conservative treatments and reduce mutilating treatments that cause significant psychological impacts to the patients.