Signal Processing for Early Warning Arrhythmia Detection and Survival Prediction for Clinical Decision

According to the British Heart Foundation, UK, there is a population of around 7 million living in the UK with heart and circulatory diseases; about 25% of all the deaths in the UK are caused by cardiovascular diseases and more than 30,000 people a year suffer cardiac arrest out-of-hospital. As people all over the world, continue to live busy and stressful lives, a vast majority of people start showing cardiac arrhythmia-related symptoms which, if not treated in time may lead to a serious heart condition or even sudden cardiac death. To identify the early-warning signs in cardiac arrhythmia, methods to identify the precursors to fatal arrhythmia were developed in this research study, using a wearable kit. To enable accurate classification between arrhythmic beats, novel feature extraction algorithms using spectral components were developed. Often a fatal cardiac arrhythmia, or a serious injury, may lead to trauma and in such situations, it becomes imperative that the critical care teams have adequate information about the patient’s health status at remote location following an ambulatory response. A real-time trauma scoring algorithm was developed, and correlation and regression analyses were performed to arrive at these scores using the physiological parameters and vital signs. It was found that with appropriate feature extraction algorithms, supervised learning classifiers could identify the precursors to arrhythmia in real time and on a resource-constrained device, regardless of time and location. The trauma scoring algorithm, implemented using ICU patients’ dataset, produced values that agreed with the patients’ status and events could be logged to electronic health records using standard clinical coding systems. It could, therefore, be concluded that regardless of situation and location of an individual, fatal arrhythmia and trauma events could be identified ahead of time before reaching a state of emergency.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

University of Essex

Year dc:date.issued

2019

Author dc:creator

• Walinjkar, Amit

• QA75 Electronic computers. Computer science
• QA76 Computer software
• R Medicine (General)
• RA0421 Public health. Hygiene. Preventive Medicine
• RZ Other systems of medicine
• TK Electrical engineering. Electronics Nuclear engineering
• U Military Science (General)
• ZA4050 Electronic information resources
• ZA4450 Databases

Language dc:language

en