Differences

This shows you the differences between two versions of the page.

--- projects:workgroups:wg_meeting_oct_13_2021 [2021/11/09 04:04] – [Invited talk] vipina
+++ projects:workgroups:wg_meeting_oct_13_2021 [2021/11/09 04:46] (current) – [Recording] vipina
@@ Line 6: / Line 6: @@
 ==== Invited talk ====
-**Title: Natural Language Processing for Clinical Excellence: The State of Practices, Opportunities, and Challenges:** Social and Behavioral Determinants of Health (SBDH) are powerful drivers of future well-being of individuals, but the clinical community rarely has access to standardized tools to systematically incorporate SBDH into clinical research and decision-making. To address this, we have been creating fundamental resources to systematically identify SBDH from within health records. We incorporate a wide range data sources such as coded clinical data (ICD codes), encoded questionnaires, and annotated texts corpora, and we apply a variety of NLP and AI methods such as heuristic-based natural language inference, conventional machine learning, and contextual neural network models. At the same time, we also focus on the dissemination of our methods and collaborating with external partners to ensure the generalizability of our models across various health systems. Results for heuristic-based, deep learning and ensemble models are promising and we have successfully validated our models on external partners sites.
+**Title: Harnessing Big Data for Population Health: Advancing NLP Techniques to Extract Social-Behavioral Risk Factors from Free Text within Large Electronic Health Record Systems:** Social and Behavioral Determinants of Health (SBDH) are powerful drivers of future well-being of individuals, but the clinical community rarely has access to standardized tools to systematically incorporate SBDH into clinical research and decision-making. To address this, we have been creating fundamental resources to systematically identify SBDH from within health records. We incorporate a wide range data sources such as coded clinical data (ICD codes), encoded questionnaires, and annotated texts corpora, and we apply a variety of NLP and AI methods such as heuristic-based natural language inference, conventional machine learning, and contextual neural network models. At the same time, we also focus on the dissemination of our methods and collaborating with external partners to ensure the generalizability of our models across various health systems. Results for heuristic-based, deep learning and ensemble models are promising and we have successfully validated our models on external partners sites.
-**Presenter:** Dr. Yanshan Wang\\
+**Presenter:** Dr. Masoud Rouhizadeh\\
-Yanshan Wang, PhD, FAMIA is vice chair of Research and assistant professor within the Department of Health Information Management at the University of Pittsburgh. His research interests focus on artificial intelligence (AI), natural language processing (NLP) and machine/deep learning methodologies and applications in health care. His research goal is to leverage different dimensions of data and data-driven computational approaches to meet the needs of clinicians, researchers, patients and customers. Prior to joining Pitt, Dr. Wang was assistant professor in the Department of AI & Informatics at Mayo Clinic. Yanshan has extensive collaborative research experience with physicians, epidemiology researchers, statisticians, NLP researchers, and IT technicians. He has served as investigators for multiple extramural NIH-funded projects and intramural operational projects. He has published over 50 peer-reviewed articles in high-impact medical informatics journals (e.g., JBI, JAMIA), and conferences (e.g., AMIA Annual Symposium, AMIA summit, IEEE BIBM). Dr. Wang is also active in organizing conference workshops and shared tasks in the medical informatics community, including the international Health NLP workshops and the national NLP clinical challenge (n2c2).
+Masoud Rouhizadeh is an Assistant Professor in the University of Florida College of Pharmacy, Department of Pharmaceutical Outcomes, under the AI in the Health Sciences Initiative. The primary focus of Dr. Rouhizadeh’s research involves applying machine learning and natural language processing methods for identifying clinical concepts from unstructured text and converting them into structured data. Another major part of his research has been developing clinical ontologies and lexical resources, as well as computational models for identifying social and behavioral determinants of health. Before joining the UF, Dr. Rouhizadeh was a Faculty Instructor at Biomedical Informatics and Data Science and the Natural Language Processing lead at the Institute for Clinical and Translational Research at the Johns Hopkins University School of Medicine. Prior to JHU, he was a postdoctoral fellow at the University of Pennsylvania’s World Well-Being Project and then at the Penn Institute for Biomedical Informatics. He obtained his Master’s and Ph.D. in Computer Science and Engineering from Oregon Health and Science University and his Master’s in Human Language Technology from the University of Trento, Italy.
+==== Recording ====
+[[https://ohdsiorg.sharepoint.com/sites/Workgroup-NLPNaturalLanguageProcessing/Shared%20Documents/General/Recordings/Meeting%20in%20_General_-20211013_130637-Meeting%20Recording.mp4?web=1|Click here to watch the recorded meeting]]