Extracting association rules from liver cancer data using the FP-growth algorithm

The five-year survival rate of liver cancer is low, 14% according to the Surveillance, Epidemiology, and End Results (SEER) Program database of the National Cancer Institute from 2003 to 2007 [3]. Since in the early stages of liver cancer, patients usually do not show signs or symptoms, improving early diagnosis is essential in order to reduce morbidity and mortality rates. Association rule mining, a popular method for discovering interesting hidden relationships or patterns between variables in large databases, has demonstrated benefit when applied to cancer detection and management. To date, however, no studies have applied it to liver cancer. The objective of this study was to apply the FP-growth association algorithm to discover patterns from liver cancer data, which can hopefully be used for early detection.