A Path Towards an Automated and Inclusive Clinical Research
Clinical trials play important roles in drug development but they often suffer from expensive, inaccurate and insufficient patient recruitment. 50% of trials are delayed due to patient recruitment issues and 25% of cancer trials fail to enroll sufficient numbers of patients. Participating patients are also not a good representative of patient population, for example only 5% of trial participants are black, while 13.4% of the U.S. population is black. The disparity is greater for Hispanic or Latino heritage, where 18% of the US population are Latinx while only 1% of the trial participants belong to this ethnic identity.
In this talk, we introduce a project in collaboration between multiple organizations in Genentech on how we are using NLP to build a decision support systems for patient-trail matching with an end-to-end deep learning approach to optimize the trial recruitment process and find qualified patients of different demographics in a timely manner.
Azadeh Mobasher
Principal Data Scientist at Genentech
Azadeh Mobasher is Principal Data Scientist, working on applications of NLP in medical domain.
Azadeh works on the Commercial, Medical, Government affairs Data Science team and is interested in the impact of using NLP to support medical interactions, decision making and a more diverse and inclusive clinical research.
Prior to Genentech, Azadeh spent time at Convoy Inc. and Microsoft in various research and data science roles. She has a PhD in Operations Research from University of Houston and is a part time lecturer at Northeastern University, Silicon Valley campus.
