Synthesis and Evaluation of Acute Myocardial Ischemia Using a Dynamical ECG Model

VALVERDE, ER; ARINI P.D.; RINCÓN SOLER, AI; BONOMINI M.P.; FERNÁNDEZ BISCAY, C

Córdoba

Congreso; XXI Congreso Argentino de Bioingeniería, X Jornadas de Ingeniería Clínica; 2017

Physiological situations are useful tools to generate sets of signals with specific characteristics in time and frequency. An application of this synthetic signals involves training and improvement of algorithms  to  detect  different  pathophysiological  conditions.  To  obtain  the  best  results, synthesized  signals  need  to  be  highly  comparable  with  the  real  ones.  This  study  implements  a procedure  to  generate  synthetic  electrocardiogram  (ECG)  signals  from  patients  referred  to percutaneous  transluminal  coronary  angioplasty  (PTCA)  and  their  respective  control  registers using the ECGSYN model. In this order of ideas, the T-wave spectral variance (TSV) index was used as  a  method  to  compare  the  information  that  synthetic  signals  contain  in  relation  to  their  real counterparts, and in this way, test the reliability of the ECGSYN. Results indicate high similitude between  real  T-waves  and  those  produced  by  the  model  with  a  maximum  error  of  11%  in  the worst case. TSV indexes from real and synthetic signals reveal similar values in control and PTCA situations.  Additionally,  no  statistical  differences  (p  =  NS)  were  found  between  these  values, indicating that the ECGSYN is capable to replicate a physiological scenario with acute myocardial ischemia