Using Novel Machine Learning Methods to Personalize Strategies for Prevention of Persistent AKI after Cardiac Surgery

点击翻译按钮获取中文摘要

My long-term goal is to integrate health informatics, data mining and machine learning to improve the care for patients with, and at risk for, acute kidney injury (AKI). I am dual trained in Nephrology and Critical Care Medicine. I am already developing my skills in health informatics. This proposal presents a five-year career development plan for NIH K08 award focused on training in advanced data mining, machine learning and their applications to critical care nephrology. To that effect, I have assembled a strong mentoring team with decades of experience in mentoring, research and leadership. The outlined career development plan in conjunction with intensive mentoring and hands-on training will provide me the perfect platform to become a leading independent investigator in the field. AKI, a complex syndrome with heterogenous phenotypes, is seen in over one third of patients undergoing cardiac surgery. Though increasing severity of AKI is associated with worse outcomes, preliminary literature has shown than persistence of AKI is itself important. While transient AKI that resolves rapidly still has worse outcomes, the outcomes are much worse for patients with persistent AKI that lasts beyond 48 hours. Additional monitoring, reassessment of causes and management options is recommended once the diagnosis of persistent AKI is established. The prevention of AKI, however, remains paramount. Preventive actions, though effective, have low compliance among cardiac surgery patients. As over 40% of AKI after cardiac surgery is transient and resolves spontaneously within 48 hours, targeted application of these actions in patients at high risk for developing persistent AKI will allow for a more focused allocation of resources. Identification of distinct phenotypes among patients with persistent AKI will further allow for identification of differential responses to therapy and lead to personalization of AKI therapy. The current AKI research, however, is focused on identification and prevention of increasing severity of AKI. The objective of this study therefore is to identify and characterize patients at risk for and with persistent AKI after cardiac surgery. This will be accomplished by addressing the following two specific aims: (1) Develop digital biomarkers to predict patients at risk for persistent AKI after cardiac surgery, (2) Determine distinct clinical phenotypes among patients who develop persistent AKI after cardiac surgery. Completion of these aims will provide a structured framework to provide personalized care to patients with AKI. It will also provide me with preliminary data and experience necessary to apply for R01 application as an independent investigator leading a data science research program in critical care nephrology.