Cancer survivors are considered at increased risk of developing new primary cancers. The Surveillance, Epidemiology and End Results (SEER) Program monograph reports that about 10% of cancer survivors in the USA were diagnosed with a second or higher-order primary cancers. This paper uses new data science approaches to better understand associations and identify patterns of multiple cancer sites. We analyzed data from SEER program registries during a 42-year period 1973 to 2015 for 688,892 patients with multiple primary cancers (MPCs). Machine-learning algorithms including association rule mining, network analysis, and page rank were used to report patterns of MPCs.

Describe the new knowledge and additional skills the participant will gain after attending your presentation.: The attendees will learn how machine-learning algorithms including association rule mining, network analysis and a page rank algorithm can be used to better understand patterns and associations between multiple primary cancers (MPCs).


Elena Manilich (Presenter)
John Carroll University

Arshiya Mariam, John Carroll University
Zachary Zinda, John Carroll University
Saima Hanif, SUNY Medical Center

Presentation Materials: