PHM applications in medicine and medical implantable device

Abstract

The paper describes the adaptation of PHM models, methods and algorithms to solve prognosis problems for biomechanical systems, in particular, prognosis of Heart Failure, problems of antiarrhythmic electrical therapy optimization of CRT devices (cardiac resynchronization therapy devices). As the basis for making prognosis in biomechanical systems and biomechanical systems with implantable electronics are the multiscale physical models of heart failure and biophysics of membranes of cardiomyocytes. At the macroscopic level, the mathematical model of ECG analyzing, allowing to determine the evolution equations of predictors of heart failure is developed. Computing algorithms of RUL analog as the time parameter, expressed by the number of cardiac cycles required to achieve predicted values of predictors or heart failure boundaries, are defined and constructed. The paper also gives a rigorous justification of the thesis about the effectiveness of PHM methods in problems of prognosis of atrial fibrillations, ventricular arrhythmia, and prognosis of sudden heart death. The results of using the prognostic methods and estimations of time to reach the borders of heart failure are demonstrated using the modified portable PHM ECG recorder in 12 leads. Modified recorder contains built-recognizing automata, supported and retrains on a remote computing PHM cluster. The basic principles of interaction between remote computing a cluster with recognizing automata of ECG recorder are described. At the same time recognizing automata manage computational cores of remote cluster, in its turn the computing cores and logic modules of the cluster are determined automatic operation and the duration of the ECG recording. Thus, the task of optimizing the duration of monitoring is solved. Also, the results of clinical trials of methods and prognosis algorithms are demonstrated.