Using AI to Understand What Causes Diseases

#artificialintelligence

Health care leaders are embracing AI. But by conducting an extensive review of case studies and research literature, we've found that their AI initiatives are predominantly focused on developing algorithms that can predict a problem such as cancer in order to make diagnoses better, faster, and less expensively. Rarely, are their organizations devoting resources to AI efforts aimed at understanding why diseases occur. To intervene as effectively as possible, both kinds of algorithms are crucial. To be clear, we are not downplaying the importance of predictive analytics to help diagnose patients.


Using AI to Understand What Causes Diseases

#artificialintelligence

Health care leaders are embracing AI. But by conducting an extensive review of case studies and research literature, we've found that their AI initiatives are predominantly focused on developing algorithms that can predict a problem such as cancer in order to make diagnoses better, faster, and less expensively. Rarely, are their organizations devoting resources to AI efforts aimed at understanding why diseases occur. To intervene as effectively as possible, both kinds of algorithms are crucial. To be clear, we are not downplaying the importance of predictive analytics to help diagnose patients.


GNS Healthcare Names Leslie Hoyt Chief Operating Officer

#artificialintelligence

WIRE)--GNS Healthcare (GNS), a leading precision medicine company that applies causal machine learning technology to massive and diverse data streams to better match drugs and other health interventions to individual patients, today announced the appointment of Leslie Hoyt as Chief Operating Officer. A veteran healthcare operations leader, Hoyt, who most recently served United Health Group (NYSE:UNH) as Senior Vice President at Optum Health, will help GNS scale its operations and strategically navigate its development during a time of tremendous growth for the company. "Leslie brings an enormous passion for improving the health of individuals and the overall quality of care," said Colin Hill, CEO and Co-founder of GNS. "Her experience will be an essential asset to GNS, as more of our existing customers, as well as new customers, turn to GNS for solutions to complex problems arising from the transition to value-based care. Health plans, providers and biopharmaceutical companies are seeking deeper and more timely knowledge regarding which patients are likely to progress with disease, and which interventions will result in better outcomes for those patients at a lower total cost of care."


Rare Disease Treatments to Be Discovered by Machine Learning and Simulation Platform

#artificialintelligence

Earlier today, Alexion Pharmaceuticals was granted the licensing rights to operate GNS Healthcare's Reverse Engineering and Forward Simulation (REFS) casual machine learning and simulation platform. Alexion intends to use the platform to accelerate both the research of rare diseases and the development of novel therapies. Earlier this year, it was instrumental in finding a new target for breast cancer. In May, GNS published data in the American Association for Cancer Research's (AACR) journal Cancer Research in which they announced the discovery of novel targets, including TRIB1, which had positive implications for the treatment of triple-negative breast cancer. In June, it was announced that the company would be teaming up with Genentech to improve the capability to clarify disease mechanisms, recognize new targets, and diagnose patient populations more accurately.


Robustly Extracting Medical Knowledge from EHRs: A Case Study of Learning a Health Knowledge Graph

arXiv.org Machine Learning

Increasingly large electronic health records (EHRs) provide an opportunity to algorithmically learn medical knowledge. In one prominent example, a causal health knowledge graph could learn relationships between diseases and symptoms and then serve as a diagnostic tool to be refined with additional clinical input. Prior research has demonstrated the ability to construct such a graph from over 270,000 emergency department patient visits. In this work, we describe methods to evaluate a health knowledge graph for robustness. Moving beyond precision and recall, we analyze for which diseases and for which patients the graph is most accurate. We identify sample size and unmeasured confounders as major sources of error in the health knowledge graph. We introduce a method to leverage non-linear functions in building the causal graph to better understand existing model assumptions. Finally, to assess model generalizability, we extend to a larger set of complete patient visits within a hospital system. We conclude with a discussion on how to robustly extract medical knowledge from EHRs.