Congestive heart failure (CHF) is a chronic progressive condition that affects the pumping power of the heart muscles. As the chronic condition is associated with high prevalence, significant mortality rates and sustained health costs, efficient detection processes are in high demand from clinical practitioners and health systems.
To tackle these concerns, Dr Sebastiano Massaro, an associate professor of organisational neuroscience at the University of Surrey – collaborating with researchers at the University of Warwick and the University of Florence in Italy – used convolutional neural networks (CNN) – hierarchical neural networks that are highly effective in recognising patterns and structures in data – as part of their approach.
According to the team, their research drastically improves existing CHF detection methods typically focused on heart rate variability that, whilst effective, are time-consuming and prone to errors. Conversely, their new model uses a combination of advanced signal processing and machine learning tools on raw ECG signals, delivering 100 per cent accuracy.
“We trained and tested the CNN model on large publicly available ECG datasets featuring subjects with CHF as well as healthy, non-arrhythmic hearts,” said Dr Massaro. “Our model delivered 100 per cent accuracy: by checking just one heartbeat, we are able to detect whether or not a person has heart failure.”
He added: “Our model is also one of the first known to be able to identify the ECG’s morphological features specifically associated with the severity of the condition.”
Dr Leandro Pecchia from the University of Warwick and president at European Alliance for Medical and Biological Engineering, explains the implications of these findings: “With approximately 26 million people worldwide affected by a form of heart failure, our research presents a major advancement on the current methodology.
“Enabling clinical practitioners to access an accurate CHF detection tool can make a significant societal impact, with patients benefitting from early and more efficient diagnosis and easing pressures on NHS resources.”
The research, by Dr Massaro and colleagues Mihaela Porumb and Dr Pecchia at the University of Warwick and Ernesto Ladanza at the University of Florence, is published in the Biomedical Signal Processing and Control Journal.
The development of healthcare technologies that use artificial intelligence (AI) has become a priority for the UK government. At the start of August, it announced £250m for the creation of a National Artificial Intelligence Lab that will help the NHS more effectively treat conditions such as cancer, dementia and heart disease. Furthermore, the Government has allocated £133m to develop healthcare technologies that use artificial intelligence (AI) and gene-based therapies.
Meanwhile, universities and research institutes throughout the UK (sometimes collaborating with universities around the world) are continuously researching on the subject and developing new technologies to accompany it. One recent example was demonstrated by Oxford University researchers, who developed a new biomarker which uses AI to identify people at high risk of a fatal heart attack at least five years before it strikes.