Automated Syncope Prediction from Physiological Signals Acquired during Tilt-Table Testing using Deep Learning Approach

Syncope also known as transient loss of consciousness which caused trouble to human daily life. As the advancement of technology, real-time prediction of syncope can be done with machine learning (ML). Head-up tilt table test (HUTT) which having a lengthy procedure and might causing discomfort to weak patient during the test. Aim of this study is to develop a prediction algorithm which able to perform real time syncope prediction based on patient's physiological signal. In this study, physiological signal has been collected from 97 subjects during the HUTT procedures by clinicians. Several features have been extracted from heart rate, systolic blood pressure and diastolic blood pressure. Two methods were developed for the task of real-time syncope prediction, which are the numerical method and ML-based method (proposed method). Additionally, 5-fold cross-validation was performed to evaluate the performance of proposed method. The training of the Long Short-Term memory model utilized achieved a result of 98.73% of test accuracy, and after utilizing the trained model into the algorithm, the proposed model achieved an accuracy of 66.65%, sensitivity of 74.64% and a prediction time of 663 ± 639 seconds. Future work can be focus on utilizing different type of model and carry out external cross validation for achieving a better result. should not include author information and should not include acknowledgments, citations or discussion of related work that would make the authorship apparent. Submissions containing author identifying information may be subject to rejection without review. Upon acceptance, the author and affiliation information must be added to your paper.