Senior Design Projects

ECS193 A/B Winter & Spring 2020

Risk predictability of atrial fibrillation

Name Uma N Srivatsa MD
Affiliation School of Medicine

Project's details

Project title Risk predictability of atrial fibrillation
Background Premature atrial complexes (PAC) frequently occur in 24 hour electrocardiogram ( ECG) monitor. There are precursor to a condition called atrial fibrillation which carries a risk of stroke.
Description We identify all patients with Holter monitor between years 2010 and 2018. Two groups - those with and those without atrial fibrillation. We collect clinical, demographic data from electronic medical record. In addition, parameters from ECG monitor would be heart rate and PAC characteristics: number of PAC, morphology of PAC, normal complex to PAC interval. Using all these parameters we need to identify a machine learning algorithm to predict occurence of atrial fibrillaiton. One set up of patients will be to program algorithm, and second set of patients will be to validate.
Deliverable machine learning algorithm to identify risk of atrial fibrillation
Skill set desirable Machine learning programming skills, EMR programming skills ( Sequel)
Phone number 916 524 6267
Client time availability 30-60 min weekly or more
IP requirement Open source project
Attachment N/A