To derive a 12-lead standard ECG from 3 measured leads using a universal patient coefficient matrix and to detect the presence of acute MI from an EV index calculated for both the measured and derived ECGs.
Twenty training ECGs of varying pathology were acquired and digitized resulting in a 300x12 voltage-time data array for mathematical processing. The simplex optimization (SOP) technique was used to derive a 12x3 universal patient coefficient matrix from the 20 case ECG training set. A different set of 55 test cases, including 37 normal and 18 acute infarction ECGs, were similarly acquired and digitized, from which leads I, aVF, and V2 were chosen as the measured 3 lead-vector basis factor space. The SOP coefficient 12x3 matrix was then multiplied by the [I, aVF, V2} measured lead-vector 3x300 matrix yielding the derived 12-lead ECG 300x12 matrix. The 55 measured and derived test case ECGs were graphically compared for diagnostic and morphologic correlation.
All 55 test case ECGs were predicted correctly. No significant morphologic or diagnostic changes were noted in the derived ECGs. Significant differences between normal and acute MI were detected at EV3% (p < 0.05) for both measured and derived ECGs and the EV index predicted pathology in all cases correctly. The reduction of the measured 12-lead ECG data set to 3 leads allowed the display of a vector plot of the movement of the electrical forces resulting in a 3-dimensioanl spatial ECG curve.
A universal patient coefficient matrix has been derived to allow 12-lead standard ECG derivations from 3 measured leads acquired using the SOP technique. This study also demonstrated that an EV index may differentiate normal from acute MI pathology.
Using this new technology, it is now possible to perform instantaneous, real-time, point of service, cost-efficient 3-lead rhythm processing using bed-side cardiac monitoring systems to produce a derived 12-lead ECG. Continuous monitoring of the EV index provides a dynamic electrical marker for acute MI.
David Schreck, None.