Name
Detecting the Undetected: Machine Learning Tests for Chronic Kidney Disease
Description
This presentation explores a random forest machine learning model developed in R to detect undiagnosed chronic kidney disease using Medicare claims data. Trained on diagnosis history, the model identifies patients at risk, enhancing early detection and intervention. By leveraging predictive analytics, this approach aims to improve patient outcomes and healthcare efficiency. The session will discuss model development, validation, and potential applications in clinical settings.
Presentation Learning Objectives
1) Understand Open-Source Software and Its Role in Data Integration and Predictive Modeling. Attendees will learn what open-source software is and how it facilitates data merging and machine learning model development. This will be demonstrated through an overview of R statistical software, its capabilities in handling electronic health records (EHR) and Medicare claims data, and its application in building predictive models.
2) Learn the Process of Developing a Predictive Model for Chronic Kidney Disease (CKD). The presentation will walk attendees through the key steps of developing a machine learning model, from data preprocessing and feature selection to training and validation. By exploring the random forest algorithm used in this study, attendees will gain a practical understanding of how predictive models are built and evaluated for accuracy.
3) Recognize the Financial and Holistic Benefits of Early CKD Detection. Attendees will learn how early detection of CKD through predictive modeling can reduce healthcare costs, prevent disease progression to end-stage renal disease (ESRD), and improve patient outcomes. Real-world implications, including cost savings for healthcare systems and better quality of life for patients, will be discussed to highlight the value of proactive intervention.
2) Learn the Process of Developing a Predictive Model for Chronic Kidney Disease (CKD). The presentation will walk attendees through the key steps of developing a machine learning model, from data preprocessing and feature selection to training and validation. By exploring the random forest algorithm used in this study, attendees will gain a practical understanding of how predictive models are built and evaluated for accuracy.
3) Recognize the Financial and Holistic Benefits of Early CKD Detection. Attendees will learn how early detection of CKD through predictive modeling can reduce healthcare costs, prevent disease progression to end-stage renal disease (ESRD), and improve patient outcomes. Real-world implications, including cost savings for healthcare systems and better quality of life for patients, will be discussed to highlight the value of proactive intervention.
Speaker Name
Raymond W Tupling Jr
Speaker Credentials
MSc
Speaker Title
Data Scientist Supervisor
Speaker Organization
Mountain Pacific Quality Health