CE 18.13

Title: 

ARtPM: Article Retrieval for Precision Medicine

Author(s): 

Lowell Milliken

Oral Defence Date: 

Wednesday, December 5, 2018 - 15:00

Location: 

TH 434

Committee: 

Asst Prof Kulkarni and Prof. Ilmi Yoon

Abstract: 

Precision medicine is an emerging patient care model that customizes treatment based on the specifics of their condition and background at a level that would have been unreasonable in earlier times. The genetic information about a patient's disease, such as the gene and mutation that may be contributing to the illness, can be used to create treatment plans targeting the genes. The difficulty is that, as a consequence of the ongoing expansion of medical knowledge, there are more published works fitting a general profile of a patient's condition than can readily be searched for those fitting the specific patient. The information retrieval task is to find these specific articles given disease, genetic and demographic information about a patient. We have designed and developed, Article Retrieval for Precision Medicine (ARtPM), an end-to-end article retrieval system that uses multi-staged approach to identify relevant articles for the given particular medical case. Thorough empirical evaluation shows that ARtPM performs significantly better than multiple baselines, including PubMed search.

Keywords: 

precision medicine, information retrieval, learning to rank

Copyright: 

Lowell Milliken