CAD patient classification using MIMIC-II

Document Type

Conference Article

Publication Title

Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST


With availability of large volume of collected data from healthcare centers and significant improvement in computation power, evidence based learning is helping in building robust disease diagnostic models. In this work MIMIC-II database, consisting of physiologic waveforms and clinical Information about ICU patients, is used for patient classification, taking Coronary Artery Disease (CAD) as a use case. A learning algorithm (wavelet transform + SVM) is trained and evaluated for CAD patient segregation with 89% accuracy on ICD-9 labeled MIMIC-II Photoplethysmogram (PPG) signals. Due to the noisy nature of machine collected MIMIC-II ICU data, the same SVM model was validated on a local hospital dataset containing doctor labeled PPG signals resulting a 5% accuracy gain. This work is the first attempt of CAD patient classification onMIMICII, using heart rates from easily obtainable PPG signal suitable in mobile/wearable setting.

First Page


Last Page




Publication Date


This document is currently not available here.