Agenda
1) Presentation - Dr. Daniel G. Smith, Emory University
Abstract: Implementing an NLP process into production is an often challenging task, with considerable requirements for accuracy prior to being relied upon in a clinical workflow. However, including other local members of the clinical and research community in the process at each stage of model selection and development could provide considerable advantages. At Winship Cancer Institute, we’re starting to bring NLP to our research abstractors, as well as our tumor registrars, so that they can provide critical feedback on dictionary curation, NER label verification, and final accuracy of the models in development. We seek feedback from the NLP working group regarding our model of the current process, as well as discuss your experiences in bringing clinical and research staff into the NLP process.
2) Updates on the progress of ongoing studies
3) NLP book chapter and other OKRs