Electrocardiographic (ECG) recordings are one of the most employed tools for diagnosing, monitoring and screening for cardiac pathologies in the population. The ECG offers a rich insight into the electrical activity patterns of the myocardium, given that its morphology is affected by a large number of conditioning factors, including specific pathologies such as myocardial infarction as well as a large array of comorbidities such as genetic or environmental factors. Different factors can have very apparent or subtle effects on the morphologies of the different parts of the cardiac cycle represented by the ECG, such as abnormally large voltage on the P or QRS waves in the case of hypertrophic patients.
The objective of this work is to develop computational phenotypes of patients with hypertrophic cardiomyopathy from ECG recordings. This will be performed by using multiple kernel learning (MKL), an unsupervised machine learning algorithm that combines different pools of input data, finding similarities and differences inherent to the data in an agnostic manner. MKL produces low-dimensional output spaces that will be used to characterize phenotypic differences that allow for a better understanding of the link between genetic factors and long-term outcomes.
 Lin, Y. Y., Liu, T. L., & Fuh, C. S. (2010). Multiple kernel learning for dimensionality reduction. IEEE Transactions on Pattern Analysis and Machine Intelligence, 33(6), 1147-1160.
 Sanchez-Martinez, S., Duchateau, N., Erdei, T., Fraser, A. G., Bijnens, B. H., & Piella, G. (2017). Characterization of myocardial motion patterns by unsupervised multiple kernel learning. Medical image analysis, 35, 70-82.
 Rogers, A. J., Selvalingam, A., Alhusseini, M. I., Krummen, D. E., Corrado, C., Abuzaid, F., ... & Narayan, S. M. (2020). Machine Learned Cellular Phenotypes Predict Outcome in Ischemic Cardiomyopathy. Circulation Research.
 Jimenez-Perez, G., Alcaine, A., & Camara, O. (2021). Delineation of the electrocardiogram with a mixed-quality-annotations dataset using convolutional neural networks. Scientific reports, 11(1), 1-11.