Our goals are to (1) help ensure safe and reliable performance of radiology-related AI applications and (2) develop AI applications that improve the quality and effectiveness of imaging-based healthcare.
Ensuring safe and reliable performance of AI applications
We are focused on developing automated methods for systematically evaluating and monitoring the performance of AI applications in the clinical environment, both before and after deployment. Through rigorous, thorough, and objective evaluation in the laboratory setting, we seek to predict AI model performance in the real world, including performance capability, failure modes, and performance degradation due to inherent variability in the clinical environment. Through intelligent monitoring techniques, we seek to minimize the reliance on healthcare professionals for quality assurance tasks.
AI applications that improve quality and effectiveness
Applications we develop improve quality and effectiveness in imaging-based healthcare primarily through performance assessment methods that enable feedback and improvement. This includes applications that evaluate and monitor other AI applications used in the clinical environment, as well as those that assess the quality of clinical performance in ways that are difficult to accomplish using traditional analytical methods.
To facilitate the transformation of healthcare systems to high reliability organizations through the use of radiology-related AI.
We are inquisitive, creative, and eager to innovate. We love learning the best methods from the brightest minds as we enthusiastically tackle important problems. As perpetual students, we do our homework. We learn from all available sources: the literature, our colleagues, peers, and even unrelated fields. We are especially committed to fully understanding the user experience and observing and listening to users to discover new challenges.
We generously share ideas, often through passionate but constructive dialogue. Our team thrives on diversity of background and thought, and welcomes all thoughtful ideas—even the unconventional ones. We embrace visitors who are committed to the same values. While we encourage sharing, we are also careful to give credit where it is due and to not take someone else’s ideas without permission and attribution. We uphold and maintain integrity and transparency in our research practices and daily work.
We are committed to having a worldwide impact that makes a difference for people and patients through the broad application and dissemination of our methods. We are cognizant of the stewardship with which our supporters have entrusted us. We work diligently and hold ourselves accountable for achieving tangible results and producing high quality outputs. We have a bias for action and regularly check ourselves to ensure we are making progress toward our goals.
