Cardiovascular diseases (CVD) are associated with high healthcare costs,as well as are a leading cause of mortality and hospitalizations. One of CVDs is a heart failure which may be associated with dyssynchrony of contraction of right and left ventricle. Chance for group of patients whose pharmacotherapy is not enough is cardiac resynchronisation therapy (CRT). Effectiveness of CRT has been proven in various multicenter clinical studies. The challenge limiting CRT usage is it relative low effectiveness - with significant group of patients that do not respond to this method of therapy. The device itself does not always show the true level of stimulation during interrogation; then invalid functioning is often not detected, which presents a real danger to patient's health and life. The main challenge for today's researchers is to develop new technologies, which may help to improve diagnosis of CVD, thereby reducing healthcare costs and quality of patients' lives. Smart computed systems of ECG analysis and interpretation offer new capabilities for the diagnosis and management of patients with CRT. Several reports with intelligent machine-based learning algorithms have been published, in which achieved very positive results in detecting various ECG abnormalities. Aim of our study is to show utility of ECG interpretation software in patients with CRT to assess the CRT response using Cardiomatics system.
The study is an investigator-initiated, single centre, prospective observational trial. The study will be carried out in university hospital on electrocardiology ward. The study will consist of two independent groups of patients whose ECG will be collected using the standard 12-lead ECG or 24-hour Holter monitoring. The study groups will be as follows: cardiac resynchronization therapy (CRT) recipients with pacemaker or defibrillation function, patients after cardiac implantable electronic devices (CIED) such as : cardiac pacemaker, patients with implantable cardioverter defibrillator (ICD) with indications for periodic heart stimulation. Approval for all study groups was obtained from institutional review board. In patients with already implanted device signal will be recorded in pacing mode and standby mode. What is more, in patients with CRT-D/CRT-P signal will be registered with different configurations of stimulation (no stimulation, right ventricle pacing, left ventricle pacing, biventricular pacing) and by stimulation different regions of left ventricle. Patients medical history will be acquired : comorbidities, qualification for device implantation, and other examinations at that time. In selected patients with typically good response for CRT, ECG signal will be registered with Holter method. Based on collected ECG, the correlation between clinical data parameters predicting good therapy response will be determined. ECG platform. The analysing system for arrhythmia detection consists of cloud-based software platform. The captured electrocardiographic signal uploaded to the platform is analysed using deep neural networks algorithms. The software allows the ECG standard report visualisation of signal and analysis of acquired data in terms of CRT sufficiency. The platform is a medical device certified in the European Union.
Study Type
OBSERVATIONAL
Enrollment
547
1st Department of Cariology of Medcial University of Warsaw
Warsaw, Masovian Voivodeship, Poland
Number of correctly assessed ECG signals by the automatic recognition of resynchronization in CRT-mediated therapy.
Evaluating the effectiveness of CRT therapy based on the record from an implantable device Assessment of the rationale for the use of machine based learning algorithms in detecting ECG abnormalities to determine which clinical conditions have impact on long-term effectiveness of cardiac resynchronization therapy using both standard 12-lead ECG and 24-hour Holter monitoring . The study might identify which clinical parameters in patients with CRT indicate the most benefit and the least benefit from CRT. It is planned to reach 99% sensitivity of automatized recognizing resynchronization in CRT-mediated therapy
Time frame: 14 months
Correctly recognized ECG signals after adding each cycle of 20 new ECG recordings from patients with electrical heart function disturbances.
To achieve this goal we will collect representative base of ECG recordings containing both paced rhythm in subjects undergoing therapy and those in qualification process in order to use the software to predict CRT response. The final model assumes fully automatized diagnosis of CRT-therapy response based on machine learning. Using this feature in connection with new methods of digital signal processing will constantly increase system's efficacy measured by simultaneous achievement of high test specificity and sensitivity. Increase by 1% of test sensitivity withholding high specificity after adding each cycle of 20 new ECG recordings from patients with electrical heart function disturbances is planned.
Time frame: 7 months
Number of registered ECG signals from patients holding a CIED.
Creation of an database of ECG and Holter monitoring acquired signal from patients with cardiac implantable electronic devices (CIED). Reaching more than 258 ECG recordings in the database to qualify and discriminate signal patterns that can be qualified as certain arrhythmia.
Time frame: 14 months
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.