TY - JOUR
T1 - Automated detection of atrial fibrillation using RR intervals and multivariate-based classification
AU - Kennedy, Alan
AU - Finlay, Dewar D
AU - Guldenring, Daniel
AU - Bond, Raymond R
AU - Moran, Kieran
AU - McLaughlin, James
PY - 2016
Y1 - 2016
N2 - Automated detection of AF from the electrocardiogram (ECG) still remains a challenge. In this study, we investigated two multivariate-based classification techniques, Random Forests (RF) and k-nearest neighbor (k-nn), for improved automated detection of AF from the ECG. We have compiled a new database from ECG data taken from existing sources. R-R intervals were then analyzed using four previously described R-R irregularity measurements: (1) the coefficient of sample entropy (CoSEn), (2) the coefficient of variance (CV), (3) root mean square of the successive differences (RMSSD), and (4) median absolute deviation (MAD). Using outputs from all four R-R irregularity measurements, RF and k-nn models were trained. RF classification improved AF detection over CoSEn with overall specificity of 80.1% vs. 98.3% and positive predictive value of 51.8% vs. 92.1% with a reduction in sensitivity, 97.6% vs. 92.8%. k-nn also improved specificity and PPV over CoSEn; however, the sensitivity of this approach was considerably reduced (68.0%).
AB - Automated detection of AF from the electrocardiogram (ECG) still remains a challenge. In this study, we investigated two multivariate-based classification techniques, Random Forests (RF) and k-nearest neighbor (k-nn), for improved automated detection of AF from the ECG. We have compiled a new database from ECG data taken from existing sources. R-R intervals were then analyzed using four previously described R-R irregularity measurements: (1) the coefficient of sample entropy (CoSEn), (2) the coefficient of variance (CV), (3) root mean square of the successive differences (RMSSD), and (4) median absolute deviation (MAD). Using outputs from all four R-R irregularity measurements, RF and k-nn models were trained. RF classification improved AF detection over CoSEn with overall specificity of 80.1% vs. 98.3% and positive predictive value of 51.8% vs. 92.1% with a reduction in sensitivity, 97.6% vs. 92.8%. k-nn also improved specificity and PPV over CoSEn; however, the sensitivity of this approach was considerably reduced (68.0%).
KW - Atrial fibrillation
KW - R-R intervals
KW - Algorithms
UR - http://www.scopus.com/inward/record.url?eid=2-s2.0-84994097144&partnerID=MN8TOARS
U2 - 10.1016/j.jelectrocard.2016.07.033
DO - 10.1016/j.jelectrocard.2016.07.033
M3 - Article
SN - 0022-0736
VL - 49
SP - 871
EP - 876
JO - Journal of Electrocardiology
JF - Journal of Electrocardiology
IS - 6
ER -