**Authors and Affiliations**
Dai-Yin Lu1,2,3, Moumita Bhattacharya4, Joseph E. Marine5, Jeffrey E. Olgin2, Stefan L. Zimmerman6, Theodore P. Abraham1,2, Hagit Shatkay4, M. Roselle Abraham1,2
1Hypertrophic Cardiomyopathy Center of Excellence, Johns Hopkins University, Baltimore, MD, USA
2Hypertrophic Cardiomyopathy Center of Excellence, Division of Cardiology, University of California San Francisco, San Francisco, CA, USA
3Department of Medicine, University of Utah, Salt Lake City, UT, USA
4Computational Biomedicine Lab, Computer Sciences, University of Delaware, Newark, DE, USA
5Department of Cardiology, Johns Hopkins University, Baltimore, MD, USA
6Department of Radiology, Johns Hopkins University, Baltimore, MD, USA

**Background**
Atrial fibrillation (AF) in hypertrophic cardiomyopathy (HCM) is associated with high risk for stroke even in patients with low CHA2DS2-VASc scores. Hence, there is need to understand AF pathophysiology and predict AF in HCM. In this retrospective study, we develop/apply a data-driven, machine learning-based method to identify AF cases and clinical features associated with AF, using electronic health record (EHR) data.

**Methods**
HCM patients with documented paroxysmal/persistent/permanent AF (n=191) were considered AF cases, and the remaining patients in sinus rhythm (n=640) were tagged as No-AF. We evaluated 93 clinical variables; the most informative variables useful for distinguishing AF from No-AF cases were selected based on the 2-sample t-test and information gain criterion.

**Results**
We identified 18 highly informative clinical variables that are positively associated (11) and negatively correlated (7) with AF in HCM. Next, patient records were represented via these 18 variables. Data imbalance resulting from the relatively low number of AF cases was addressed via a combination of over- and under-sampling strategies. We trained and tested multiple classifiers under this sampling approach, showing effective classification. Specifically, an ensemble of logistic regression and naïve Bayes classifiers, trained based on the 18 variables and corrected for data imbalance, proved most effective for separating AF from No-AF cases (sensitivity=0.74, specificity=0.72, C-index=0.80).

**Conclusion**
Our model (HCM-AF-Risk Model) is the first machine learning-based method for identification of AF cases in HCM. This model demonstrates good performance, and suggests that AF is associated with a more severe cardiac HCM phenotype. [Full paper]



**Authors and Affiliations**
Dai-Yin Lu^1,2,3^, Moumita Bhattacharya^4^, Joseph E. Marine^5^, Jeffrey E. Olgin^2^, Stefan L. Zimmerman^6^, Theodore P. Abraham^1,2^, Hagit Shatkay^4^, M. Roselle Abraham^1,2^
^1^Hypertrophic Cardiomyopathy Center of Excellence, Johns Hopkins University, Baltimore, MD, USA
^2^Hypertrophic Cardiomyopathy Center of Excellence, Division of Cardiology, University of California San Francisco, San Francisco, CA, USA
^3^Department of Medicine, University of Utah, Salt Lake City, UT, USA
^4^Computational Biomedicine Lab, Computer Sciences, University of Delaware, Newark, DE, USA
^5^Department of Cardiology, Johns Hopkins University, Baltimore, MD, USA
^6^Department of Radiology, Johns Hopkins University, Baltimore, MD, USA

**Background**
Atrial fibrillation (AF) in hypertrophic cardiomyopathy (HCM) is associated with high risk for stroke even in patients with low CHA2DS2-VASc scores. Hence, there is need to understand AF pathophysiology and predict AF in HCM. In this retrospective study, we develop/apply a data-driven, machine learning-based method to identify AF cases and clinical features associated with AF, using electronic health record (EHR) data.

**Methods**
HCM patients with documented paroxysmal/persistent/permanent AF _(n=191_) were considered AF cases, and the remaining patients in sinus rhythm (_n=640_) were tagged as No-AF. We evaluated _93_ clinical variables; the most informative variables useful for distinguishing AF from No-AF cases were selected based on the 2-sample t-test and information gain criterion.

**Results**
We identified 18 highly informative clinical variables that are positively associated (11) and negatively correlated (7) with AF in HCM. Next, patient records were represented via these 18 variables. Data imbalance resulting from the relatively low number of AF cases was addressed via a combination of over- and under-sampling strategies. We trained and tested multiple classifiers under this sampling approach, showing effective classification. Specifically, an ensemble of logistic regression and naïve Bayes classifiers, trained based on the 18 variables and corrected for data imbalance, proved most effective for separating AF from No-AF cases (_sensitivity=0.74_, _specificity=0.72, C-index=0.80_).

**Conclusion**
Our model (_HCM-AF-Risk Model_) is the first machine learning-based method for identification of AF cases in HCM. This model demonstrates good performance, and suggests that AF is associated with a more severe cardiac HCM phenotype. [[Full paper]](https://www.cjcopen.ca/article/S2589-790X(21)00031-7/fulltext)



HCM-AF-Risk Model to Identify Cases and Predictors of Atrial Fibrillation in Hypertrophic Cardiomyopathy