How will you revolutionize health?
The 2020 National Health Symposium will feature a technical poster session showcasing the application of artificial intelligence and machine learning in solving critical health challenges across the healthcare ecosystem. The posters will be displayed throughout the event, with formal presentations during a designated session. Presenters will also have the option to include their poster materials in the published event proceedings.
Please submit all poster abstracts to JHU-APL-NationalHealth@jhuapl.edu for consideration. Posters may detail projects in any stage of development, and will be selected based on relevance to the Symposium’s overall themes and session topics (see below).
DEADLINE EXTENDED: The deadline for submitting abstracts is now January 31, 2020.
Submit Your AbstractKey Dates
Sessions
Unlocking the Power of AI for Healthcare
To realize the potential of AI in healthcare, we must achieve the right balance between human and machine intelligence. Machines can help us perceive. The proliferation of health analytics has illustrated that machines may outperform humans in recognizing complex patterns in data, such as laboratory data and images. Machines can also help us decide. As people, we are limited in the number of possibilities we can consider. Machines aid us by evaluating a vast space of possible actions, such as complex treatment decisions, and recommending the one that best achieves our goals. In some cases, machines can act to perform health-related tasks like image-guided robotic interventions with improved speed and accuracy. To have impact, all machines must team. Using machine learning to discover insights into our own work patterns and preferences may ultimately help us create more effective machine teammates for health professionals. How do we recognize the potential applications of AI in healthcare? Which emerging techniques in AI hold the most promise to reliably automate key aspects of healthcare workflows? What can we learn from early successes and failures? This panel will convene leaders from industry, academia and government to explore these questions and more.
Explaining the Performance of AI in Healthcare
AI applications are being developed across the healthcare continuum to enable better perception, decisions, and actions. Clinical operational applications include disease prevention, disease detection, case triage, patient treatment, and disease and health monitoring. A variety of AI tools can be applied across these areas, including those that focus on patient-generated data analysis (wearables and biosensors), image analysis, and electronic health record-embedded applications. Many of these solutions are focused on the clinician, but may also enable care without traditional healthcare resources. How should we measure the integration of AI and its impact? What considerations are ‘special’ to AI —opportunities and risks, safety and security, explainability, application validation, and information assurance? This session will explore how these tools can be most impactful, who uses them, and how their performance can be studied and measured.
Ensuring Responsible Implementation of AI in Healthcare
AI is poised to become a transformational force in healthcare, but there are obstacles to implementing it broadly. Developing the technology itself is challenging. However, many implementation obstacles tend to be non-technical and are centered on creating organizational structures to ensure AI is used in compliance with policy and broader societal expectations, such as trust and accountability. When can data be shared without compromising privacy? Will AI introduce errors and biases that degrade the standard of care or cause inconsistent application across populations? If something goes wrong, who’s to blame? To realize AI’s potential in clinical practice, we must address the challenges during development and before implementation. This panel will focus on practical mechanisms for overcoming implementation obstacles, from policy and organizational solutions to technical mitigations.