METHOD FOR SMART DETECTION OF R WAVE PEAK ON THE ECG ACQUIRED INSIDE MRI SCANNER
Publication Date: 2014-Apr-04
The IP.com Prior Art Database
The invention proposes a technique to detect R wave peak from an electrocardiogram (ECG). When a patient is undergoing regular MR scan the R wave peak becomes corrupted with magnetohydrodynamic (MHD). R wave peak which are corrupted with MHD artifacts is detected. The technique uses wavelet transform as basis and threshold is applied to detect the R wave peak based on signal characteristics. ECG data is first preprocessed to remove baseline drift and noise. The contaminated ECG signals are decomposed into various frequency sub-bands using db4 wavelet. For each level, detail signals are reconstructed using low-pass and high-pass reconstruction filters with up-sampling. After reconstruction of the detail signals, ECG reference signal is obtained which results in the frequency sub-band. Such ECG reference signal is used to determine the R wave peak. The ECG reference signal is then differentiated twice to obtain a cumulative differentiated signal. The threshold is applied to the cumulative differentiated signal.
FIELD OF INVENTION
The invention generally relates to a magnetic resonance imaging (MRI) scanner and more particularly to a technique for detection of R wave peak on an electrocardiogram (ECG) signal acquired inside MRI scanner.
BACKGROUND OF THE INVENTION
Generally, cardiac magnetic resonance imaging (MRI) requires synchronization of electrocardiogram (ECG) signal with image acquisition. Complete acquisition is not acquired in one heart cycle. Hence, successive acquisitions are required to combine with a cardiac phase motion. There are various triggering techniques. Such techniques include finger plethysmography, carotid pulse tracings and ECG. ECG is the most widely used trigger. ECG is preferred as electrical activity of heart bears a constant relationship to mechanical activity of the heart. ECG is also preferred as R wave precedes mechanical systole. Obtaining ECG with a patient in a bore of a MRI scanner is difficult due to large static and dynamic magnetic fields present in the MRI scanner. Patient is subjected to a large static magnetic field inside the MRI scanner regardless of whether or not the MRI scanner is imaging. In presence of static magnetic field, electrical voltages are generated within a body. The electrical voltages are generated within a body due to low potentials which are the result of blood moving within the patient. Such artifacts are known as magnetohydrodynamic (MHD) artifacts or flow artifacts. ECG is a recording of detected voltages. The induced flow potentials alter ECG trace and results in false triggering for imaging acquisition. Effect of MHD is directly proportional to field strength. The impact is more with high field MR scanners.
A conventional technique relates to an algorithm based on neural network. The algorithm based on neural network is designed to eliminate the magnetohydrodynamic (MHD) artifact for ECG signal acquired during MR imaging.
Another conventional technique relates to a method for extracting magnetohydrodynamic signal from ECG. The extracted magnetohydrodynamic blood flow potential is meaningful due to relationship to blood flow. Removal of the magnetohydrodynamic voltages from ECG improves gating and provides diagnostically useful ECGs.
However the conventional techniques do not reduce magnetohydrodynamic artifacts during magnetic resonance imaging of high field MR scanners.
Hence there exists a need in the art for a technique that reduces magnetohydrodynamic artifacts during magnetic resonance imaging of high field MR scanners.
BRIEF DESCRIPTION OF THE INVENTION
The invention proposes a technique to detect R wave peak from an electrocardiogram (ECG). When a patient is undergoing regular MR scan, the R wave peak becomes corrupted with magnetohydrodynamic (MHD). R wave peak which are corrupted with MHD artifacts is detected. The technique uses wavelet transform as basis and threshold is applied to detect the...