• Solano, J.
• Calderon, A. J.
• Marin, E. Reyes
• Sandoval, H.
• Paguada, S.
• Funes, K.
• Palma-Mendoza, R. J.
• Hernandez, O.

Abstract

<jats:title>Abstract</jats:title><jats:sec><jats:title>Funding Acknowledgements</jats:title><jats:p>Type of funding sources: Public Institution(s). Main funding source(s): Universidad Nacional Autonoma de Honduras.</jats:p></jats:sec><jats:sec><jats:title>Introduction</jats:title><jats:p>Cardiovascular diseases constitute the majority of Noncommunicable Diseases deaths worldwide. In Honduras, cardiovascular diseases represent the fifth cause of death between 45 to 49 years old; 20% of emergency room visits are due to cerebrovascular events, heart failure and acute myocardial infarction.</jats:p></jats:sec><jats:sec><jats:title>Purpose</jats:title><jats:p>Create a low-cost ECG and digital platform that provides portability and connectivity with mobile devices and automated diagnosis to support clinicians in primary care.</jats:p></jats:sec><jats:sec><jats:title>Methods</jats:title><jats:p>A low-cost 3-lead ECG (Handy EKG) was developed under the technological project approach, which has two general stages: (i) platform design and (ii) field validation. The hardware design was carried out by comparing microcontrollers, 3D printing the cases and beta testing with volunteers to accomplish prototype stability. Second, the mobile application and web platform were designed to host the reference database and establish live communication between the mobile phone and the machine learning process. During the validation stage, an automatic learning model was designed to allow the device to learn from each reading and provide a possible automatic diagnosis to the user. Fifty-one device readings were performed on volunteers and compared with those obtained with a conventional 12-lead ECG and interpreted by clinicians. The results were analysed by calculating central tendency and standard deviation measures, comparing the duration, amplitude and morphology of the different ECG waves, complexes and relevant segments both on the conventional and those made with the prototype.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>The device provided readings of lead 1 of a standard 12-lead ECG to monitor and diagnose bradyarrhythmia and tachyarrhythmias. In addition, it was found that it is possible to obtain leads II and III by alternating the positioning of the electrodes. 96% (49) of the readings showed a similarity in morphology, amplitude and duration of waves, segments and complexes that allowed an interpretation of the trace. In the two cases (4%) where there was a difference, the diagnosis with the conventional ECG was pathological (abnormal). However, the diagnosis with the prototype was non-pathological (normal).</jats:p></jats:sec><jats:sec><jats:title>Conclusion</jats:title><jats:p>The clinical study results indicated that the designed platform was safe, had good quality in its operation at all levels and provided results that allow diagnoses equivalent to those of a conventional ECG in most cases. However, the 3-lead capacity shows a lower capacity to detect other pathologies that require a 12-lead or 15-lead ECG. The tests also showed that the prototype is more sensitive to noise (interference) than the conventional device, which made some readings difficult, which is considered one of the immediate steps to improve. The current results open the way to developing more prototypes and conducting larger-scale studies where cardiologists and a more significant number of patients with diagnosed cardiac pathologies.</jats:p></jats:sec>

Topics

• impedance spectroscopy
• morphology
• size-exclusion chromatography
• machine learning