Exploring ontology expansions for clinical case retrieval in precision medicine

Conference Paper
Carterette, Mohammad Alsulmi and Ben . 2017
Conference Name: 
Text Retrieval Conference
Conference Date: 
Wednesday, November 15, 2017
Sponsoring Organization: 
NIST
Publication Abstract: 

The TREC Precision Medicine (PM) track is motivated by the research problem of developing effective search systems for finding treatment plans based on an individual's genetic, environmental, and lifestyle profile. The newly introduced PM track can be seen as an extension to the previous TREC Clinical Decision Support (CDS) track with more specialization on patients with cancer diseases (i.e., precise questions vs generic questions).

This year's track consists of two main search tasks with both tasks relying on a set of 30 topics. The first task focuses on retrieving biomedical articles which contain relevant treatments for each given patient, while the second task addresses the challenge of finding clinical trials which are available to diagnosed patients. The topic set for this year's challenge consists of synthetic cases describing the cancer type, the genetic variants, and some other potentially relevant history about the patient. In addition, two dierent document collections -abstracts of scientic articles and descriptions of clinical trials- are used as target collections for the first and second tasks, respectively.