A Concept for Mining Transitive Sequential Patterns from Pancreatic Cancer Patient Journeys

Stud Health Technol Inform. 2024 Aug 30:317:235-243. doi: 10.3233/SHTI240862.

Abstract

Pancreatic cancer, renowned for its aggressive nature and poor prognosis, necessitates the optimization of treatment strategies. The sequence of procedures in clinical trials is critical, such as evaluating the potential benefits of preoperative chemo-radio-therapy for pancreatic cancer. Nevertheless, we might not be aware of other temporal sequences which have an effect on therapy response or the general outcome. Extracting transitive sequential patterns from patients' medical trajectories allows researchers to identify temporal characteristics for complex diseases. We illustrate how such sequential patterns can be discovered and might be utilized in pancreatic cancer research as well as patient care.

Keywords: pancreatic cancer; sequential pattern mining; temporal pattern mining.

MeSH terms

  • Data Mining*
  • Humans
  • Pancreatic Neoplasms* / therapy