Heterogeneous signal intensity in magnetic resonance images of hypertrophied left ventricular myocardium

• Published in: Magnetic resonance imaging Vol. 7; no. 5; pp. 517 - 528
• Main Authors: Zahler, Raphael, Chelmow, David, Gore, John, Wilkens, Kenneth, Pope, Christopher, Sostman, H.Disk, Meese, Roderick, Negro-Vilar, Rosa, Herfkens, Robert J., Wackers, Frans J., Zaret, Barry L.
• Format: Journal Article
• Language: English
• Published: New York, NY Elsevier Inc 01.09.1989; Elsevier Science
• Subjects: Adult; Aged; Biological and medical sciences; Cardiology. Vascular system; Cardiomegaly - diagnosis; Cardiomegaly - pathology; Heart; Heart failure, cardiogenic pulmonary edema, cardiac enlargement; Humans; Image processing; Left ventricular hypertrophy; Magnetic Resonance Imaging; Medical sciences; Middle Aged; Myocardium - pathology; Tissue characterization; Magnetic resonance imaging; Image processing; Tissue characterization; Left ventricular hypertrophy; Human; Myocardium; Nuclear magnetic resonance imaging; Left ventricle; Hypertrophy
• ISSN: 0730-725X; 1873-5894
• DOI: 10.1016/0730-725X(89)90405-0

Left ventricular hypertrophy is associated with decreased longevity and often leads to congestive heart failure. An exploratory study of magnetic resonance imaging in human left ventricular hypertrophy was performed. First, 13 patients with left ventricular hypertrophy and 7 controls of similar ages were studied using electrocardiogramgated end-diastolic images. Visual inspection suggested that low-intensity zones were frequently found within the hypertrophied myocardium. To verify this observation, the images were processed with semi-automatic edge detection and a derivative-based tissue characterization algorithm, yielding tissue heterogeneity indices (THI-A and THI-V) which objectively measured the low-intensity zones. THI-A and THI-V were both significantly greater in left ventricular hypertrophy patients than in controls (THI-A: 0.111 vs 0.038, p = 0.009). THI was also significantly correlated with duration of disease and electrocardiographic abnormalities. To validate these initial findings prospectively, the same quantitative analysis was applied to magnetic resonance images of an additional 20 left ventricular hypertrophy patients and 12 controls from two institutions, using different imaging systems and different acquisition parameters. Again, THI was significantly greater in patients than in controls. Analysis of end-systolic images yielded similar results. In four dogs with left ventricular hypertrophy induced by aortic banding, THI showed a statistically significant increase as left ventricular hypertrophy developed. Hypertrophied myocardium thus shows reproducible differences from normal tissue with magnetic resonance imaging; hence, quantitative magnetic resonance tissue characterization may be useful in assessing pathologic changes in LVH.