Cross-Interval Continuous Unification Model for Abnormal Heart Rhythm Detection Using ECG Signals

Electrocardiogram (ECG) signals observed from wearable sensing devices follow a cyclic and continuous pattern for detecting heart-related diseases.Signal processing for medical psychological vital observation identifies the random or irregular rhythms guided by machine learning and artificial intelligence techniques.In this article, a Cross-Interval Continuous Unification Model (CICUM) is introduced to identify such irregular rhythms in continuous observation intervals.This model is backboned by a two-layer neural network for continuity verification and signal correlation.The first layer is used to identify discontinuous sequences between fixed signal sensing intervals.The second layer is responsible for correlating the peak and lower-order signal pulses with the normal and abnormal ECG training inputs.In the first layer, the continuous time interval metric is used to identify discontinuities that are converged using identified signal iterations.In the second layer, the variations between high and low pulses are used to train the neural network for precise abnormal signal detection.Therefore, the proposed model unifies cross intervals and signal correlations between abnormal and normal sequences to detect heart-related diseases from irregular psychological signals.The CICU improves the sequence classification, detection accuracy, and continuity verification by 14.71%, 8.91%, and 11.82% respectively for the maximum sequences.