Evaluation of a Gaussian Mixture Model forgenerating synthetic ECG signals during anangioplasty procedure

RINCON SOLER ANDERSON; PEDRO ARINI; SANTIAGO CARACCIOLO; BONOMINI, MARÍA PAULA

Tenerife

Conferencia; 9th International Work-Conference on the Interplay between Natural and Artificial Computation; 2022

Mathematical models have intensively been used in the generationof synthetic electrocardiogram (ECG) signals. They are used totest and optimize different algorithms for patient monitoring and heartdisease detection. In this work, we present a Gaussian mixture model thatallows the generation of heartbeats with ischemic alterations, induced byan occlusion procedure performed in the right coronary artery. Realizationsobtained using the Gaussian mixture model were compared againstreal signals using cross-correlation, obtaining average values equal to94.2%. Confidence intervals, at 99% level, calculated using ST valuesfrom the synthesized heartbeats agreed with the real ST values in 100%cases for each group (ECG Lead and ST-level). In this sense, the modelproposed allows the synthesis of heartbeats with complex morphologiesthat cannot be obtained with more traditional models.