ECG analysis for heart diagnostics
Simula Research Laboratory has developed numerical models that can be used to predict the effect on the surface ECG-signals from different ischemia patterns in the hearth.
The Inverse Problems group at Simula have initiated a project with the aim is to use such model in an inverse mode for predicting the ischemia patterns in the hearth based upon measured surface multi-channel ECG signals. However, the measurements need to be done while the patient is exercising, causing very noisy and sometimes erroneous data.
SINTEF was engaged in order to develop new signal processing and pattern recognition algorithms that would allow accurate and robust measurements from such noisy and unreliable multi-channel ECG data.
If you are interested in more information please contact
Tom Kavli
.
|
|
|
a) Noisy signals as recorded
b) Overlay of multiple hearth beats after cleaning of the signals. The signal cleaning includes filtering of 50 Hz, accurate identification of beat times, compensating for baseline drifts, and removal of deviating beat signals. |
|
|
|
Measurements are done for each of the cleaned and accepted beat signals. Here measured ST-shifts with confidence intervals are shown.
|
