Abstract

In this talk I will overview the Medical Information Mart for Intensive Care (MIMIC) database, the first publicly available critical care database. I will discuss our motivations in distributing MIMIC to the world, the challenges in publicly releasing medical data, and the impact it has had on the community. I will cover approaches we have created for de-identifying patient clinical data, images, and free-text clinical notes. Notably, I will describe the hybrid rule-based and neural network approach we developed for de-identifying clinical notes which achieved a sensitivity of 99.92% on a real world corpus of clinical notes. I’ll conclude with my thoughts regarding the future of healthcare data modeling, including the need for novel data structures which ensure data quality and utility for broad research use.

Speaker’s bio

Dr. Johnson is a Scientist at the Hospital for Sick Children. He received his Bachelor of Biomedical and Electrical Engineering at McMaster University and successfully read for a DPhil at the University of Oxford. Dr. Johnson has worked in machine learning and healthcare for over a decade. Dr. Johnson has strived to overcome barriers to data access, and his work on the MIMIC databases demonstrates the immense potential of publicly available healthcare data. His research focuses on enabling research through the development of new database structures tailored for healthcare, algorithms for deidentification of rich clinical data, and new machine learning methodologies for medical data.

Abstract

Violence has been identified as a public health crisis. As a counter to violence, street outreach work is a prevention model in which organizations hire residents with strong relationships and local expertise to mediate violent conflicts in their communities. In this talk, I will present the results of a two-year co-design study between a street outreach organization and academic researchers in which we collaboratively designed, built, and deployed a mobile application, Street Peace, to support street outreach workers (SOWs). Used by three different organizations in Chicago for three months, results suggest that Street Peace supported SOWs’ transformative justice practices to build a counter-structure to traditional policing, which is historically oppressive to Black communities. The SOWs used the app to mediate potentially violent conflicts without police involvement, build community through in-person events, and extend their communities of care through positive stories and narratives that countered harmful stereotypes about Black criminality. By affording SOWs more agency over their communication with each other, the app enabled SOWs to connect their strengths and scale their existing practices that combat structural oppression and prefigure liberatory futures.

Speaker’s bio

Sheena Erete, Ph.D. is a researcher and technology designer, who has studied the role of technology in addressing various social issues in Chicago for over a decade.

Dr. Erete is an assistant professor in the College of Computing and Digital Media at DePaul University and co-director of the Technology for Social Good Research and Design Lab. With a background in computer and social science, she is an expert in understanding how to situate and evaluate the use of technologies in resource-constrained communities to address issues such as violence, civic engagement, health and STEM education.

Dr. Erete has a PhD in Technology and Social Behavior (a joint degree in computer science and communication) from Northwestern University, a Master’s in Computer Science from Georgia Tech, and Bachelor degrees in Computer Science and Mathematics from Spelman College. More can be found at www.sheenaerete.com

Abstract

Because of lifestyle change and aging population, the world is facing serious healthcare challenges in terms of ever-increasing disease prevalence and medical expense. There are huge opportunities to leverage artificial intelligence technologies to transform how healthcare services are delivered, by leveraging disease prediction, smart triage, medical imaging analysis, CDSS and patient engagement models. Given the complexity of healthcare data and richness of medical knowledge, healthcare domain is one of the best areas to combine knowledge-driven and data driven AI models to solve real world challenging problems. In this talk, several concrete healthcare use cases will be shared to demonstrate how a large Medical Knowledge Graph, with 1.5 million terms, 15million relationships and 30 million medical evidence, can be combined with machine learning technologies like Deep Learning, Graph Neural Network and Reinforcement Learning etc. in real healthcare applications to transform the daily practice of physicians in the end-to-end disease management lifecycle. Future directions about how to better integrate knowledge and data-driven methods will also be discussed.

Speaker’s bio

Guotong XIE, Ph.D., Chief Healthcare Scientist of PingAn Group, Deputy General Manager of PingAn Technology, Director of PingAn Healthcare Technology Research Institute. Dr. Xie is leading the R&D of Healthcare Artificial Intelligence (AI) technologies in PingAn Group. Using the Group’s three core technologies of AI, Blockchain and cloud computing, PingAn Smart Health provides End to End disease management solutions, including smart triage and guidance, smart disease prediction, smart medical imaging screening/diagnosis, smart clinical decision support, smart patient follow-up, education, and AI-assisted drug discovery. PingAn Smart Healthcare solutions have been deployed in more than 20,000 healthcare institutions, providing 100,000+ diagnosis and treatment recommendations to 600,000 physicians every day. 270 million medical images have been analyzed to help radiologists to read X-ray and CT.

Before joining PingAn, Dr. Xie is the director of IBM Research China cognitive healthcare department and co-leading the IBM global healthcare strategy. Dr. Xie has more than 20 years in Artificial Intelligence and applying the technologies in healthcare areas. He has published more than 150 papers in top international conferences and obtained more than 80 patents. He has obtained IBM outstanding awards three times and was granted the “Best of IBM” in 2016.

Abstract

Every informatics leader has read papers from other sources that share  similar findings as something they’re working on themselves. You can  admire them from afar… Or join them and produce collaborative work products! The session explores the challenges in setting up a collaboration, executing the work, while also exploring some of the rewards from doing this.

Speaker’s bio

David Meyer, SVP of Informatics at Nations Benefits, has more than 20 years of leadership experience in healthcare and government programs operations, quality and data science. He has led informatics and analytics organizations within physician groups, regional and national health plans, and currently leads and informatics team at Nations Benefits, one of the premier Medicare supplementary benefits companies in the nation He is a frequently invited speaker at healthcare conferences and summits focused on health services, quality and operations.

Abstract

Exploiting the petabytes of laboratory data will be the differentiator for outreach reference laboratories.  While this is a much discussed and talked about concept, few laboratories have really leveraged the power of data to assist with population health, engagement with payors, or growth initiatives,  This keynote will speak to how Sonora Quest Laboratories has leveraged its data, to showcase its innovativeness, its passion for patient care and population health, and use insights to be a true data driven organization.

Participant should get a deep understanding of the role of clinical laboratories and the important function it plays in patient care, how data can be leveraged for improving care, and how labs use data to manage their operations and grow.

Speaker’s bio

Sonny Varadan is an accomplished information services executive with a passion for innovation and agility. Mr. Varadan currently serves in the capacity of Chief Information Officer for Sonora Quest Laboratories, which is the largest integrated Joint Venture laboratory in the United States and the leading reference laboratory in the state of Arizona. He has held senior IT positions including Chief Information Officer for several large reference laboratories and health systems.