event-icon
Description

To support the assessment of response in myeloproliferative neoplasms, we created and evaluated an automated information extraction pipeline to identify myeloblast counts, cellularity, and fibrosis from bone marrow biopsy reports. The pipeline displayed good performance despite high degrees of lexical variation in the concept terms and can support determination of response in these conditions, as well as other liquid tumors.

Learning Objective: This poster should help attendees understand the elements required to assess response in myeloproliferative neoplasms, as well as efforts under way to automate extraction of these elements.

Authors:

Evan Sholle (Presenter)
Weill Cornell Medicine

Spencer Krichevsky, Weill Cornell Medicine
Sajjad Abedian, Weill Cornell Medicine
Prakash Adekkanattu, Weill Cornell Medicine
Diana Jaber, Weill Cornell Medicine
Niamh Savage, Weill Cornell Medicine
Joseph Scandura, Weill Cornell Medicine
Thomas Campion, Weill Cornell Graduate School of Medical Sciences

Presentation Materials:

Tags