Completed Project (2012-2016)

Research Team: Onur Guven (PhD thesis), Amir Eftekhar, Timothy Constandinou
Collaborators: Wilko Kindt and Giovanni Frattini
FundingTexas Instruments (formerly National Semiconductor Corporation)

Image of heart and ECG waveform

Electrocardiography (ECG) is the main diagnostic tool 
for detecting cardiac disorders
. This non-invasive procedure can provide crucial information for clinicians. According to the World Health Organization (WHO), cardiovascular related diseases (CVDs) 
are still the main cause of deaths globally. Therefore, the need to improve modern healthcare systems for the reliable diagnosis and early detection of CVDs is certainly a priority.

With the advent of medical device technology, mobile and ambulatory applications prove to be the new advancement in pre-detection of coronary heart diseases and many others. Therefore, there is an increasing demand by both professionals and patients in shifting from hospitalised care solutions to home care detection systems in order to act before heart disorders reach critical levels. However, these types of systems have to work through critical challenges such as maintaining high accuracy and removing noise artefacts.

In the current project a closed loop system approach for maintaining the ECG signal integrity has been investigated throughout. The main focus is to estimate the baseline wander, electrode offset and motion artefacts in the digital domain and to subtract from the original signal through a feedback mechanism. This feedback operation of the overall system avoids conventional high pass filtering as in low resolution ECG solutions, and eliminates low frequency distortion to the ECG signal and provides real time ECG measurements. While maintaining these requirements, computationally efficient baseline detection algorithm and a hybrid interpolation algorithm minimizes the number of operations and the power requirements of the overall system without requiring bulky computerised systems.

Relevant Publications