The effect of normalization in reducing variability in regional wall thickening

This study was performed to evaluate a new method for measuring regional left ventricular wall thickening in terms of its variability, normal range, and diagnostic sensitivity. Two-dimensional echocardiographic images in the parasternal short-axis view of 44 normal patients and 17 patients with recent myocardial infarction mere analyzed. Regional wall thickness was measured according to the centerline method at 100 chords. Wall thickening at each chord i was calculated as the change in wall thickness between end diastole and end systole at chord i, normalized by the average end-diastolic thickness of the wall segment extending from chord (i - n) to chord (i + n). Five widths of the normalization segment, defined as (2n + 1), were tested: I (baseline method), 5, 11, 15, and 21 chords. Widening the normalization segment significantly reduced point-to-point variability by up to 20% (p = 0.0001) evenly over the entire left ventricular contour. The magnitude of reduction approached a plateau at a normalization segment width of II chords. The normal mean and standard deviation for regional wall thickening were also reduced slightly but significantly in a progressive and uniform manner. The apparent severity of hypokinesis worsened by up to 0.1 +/- 0.1 SD (p = 0.006) as the normalization segment was widened. The circumferential extent of hypokinesis was not lengthened significantly until the normalization segment was 21 chords wide. The magnitude of hyperkinesis was unchanged. Normalizing over wider segment of the myocardium reduces variability in the measurement of wall thickening, widens the separation between normal patients and those with disease, and enhances the diagnostic accuracy of quantitative two-dimensional echocardiography. The optimal normalization segment width is 11 chords.