Online Journal
IF値: 0.677(2017年)→0.966(2018年)


Journal of Medical Ultrasonics

にて英文誌のFull textを閲覧することができます.


1990 - Vol.17

Vol.17 No.06

Original Article(原著)

(0586 - 0597)


Ultrasonic Tissue Characterization: Analysis of Myocardial Tissue Texture in Cardiomyopathy and Myocardial Diseases

高元 俊彦, 新田 政男, 辻林 隆, 谷口 興一, 丸茂 文昭



Second Department of Internal Medicine, Tokyo Medical & Dental University

キーワード : Ultrasonic tissue characterization, Cardiomyopathy, Myocardial disease, Cardiac amyloidosis, Fabry disease

We examined the possibility of obtaining diagnostic information using digital processing ultrasound data from cardiac patients. Six healthy subjects and 18 patients with various types of cardiomyopathy (cardiac amyloidosis: 6, hypertrophic cardiomyopathy: 6, Fabry's disease: 3, acromegaly: 2, DIDMOAD syndrome: 1) were studied. We obtained our data using 3.5 MHz linear transducer scanners with a Hewlett Packard ultrasound system (77020A). A 16 mega byte digital image processing computer (Kontron, Mipron I) was interfaced to the ultrasound system.
Either real time or viedo recorded two dimensional echocardiograms covering 4 to 6 beats of cardiac cycle were stored by the computer. Reliable and reproducible data was stored with the intention of standardizing interpretation of 2 DE images; Uniform total gain control (TGC) was set-up with equal echo-brightness of interventricular and posterior wall endocardium as a reference. Using the image processing computer, 16 gray levels were color encoded. Myocardial tissue texture in 8-10 squares of the region of interest (ROI) in each frame in both end-systolic and end-diastolic periods were analyzed using a histogram method (an analysis of the distribution of echo-intensity).
In the normal subject, myocardial gray levels changed with muscular fiber orientation or depending on myocardial contraction. More gray values were noted in the septal wall where the echo-beam hit the myocardial fiber perpendicularly than in the lateral wall where the echo-beam was parallel to the fiber. Also, more myocardial gray values were noted in diastole than in systole.
On the other hand, the myocardial gray values were not changed in cardiomyopathy related to the cardiac contraction. Fewer gray level changes were seen in cardiomyopathy patients with respect to myocardial fiber orientation. The histogram from normal control patients showed narrow bands with large skewness, while those from cardiomyopathy patients showed wide bands with slight skewness.
Speckle echoes within the myocardium were seen in all patients with cardiac amyloidosis, Fabry's disease and DIDMOAD syndrome. These patients also had a wide band and low kurtosis in their histogram pattern analysis. The distribution of the speckle echoes was diffuse in amyloidosis patients but spotty in HCM and acromegaly patients. Homogeneity of the myocardial texture was defined by the standard deviation (S. D.) of gray level distribution in a histogram; Smaller S. D. means homogeneous and is color-encoded with pink, and larger S. D. means non-homogeneous and is cotor-encoded with dark red. Myocardial texture in a normal subjects was homogeneous in most parts of ROI, while it was non-homogeneous in cardiac amyloidosis patients. The prospective diagnostic accuracy of cardiac amyloidosis using a histogram analysis was 71% of sensitivity and 80% of specificity.
In conclusion, 1) non-homogeneous myocardial texture (speckle echoes) were most specific in the diagnosis of amyloidosis, 2) little myocardial gray value changes between systole and diastole suggested the substance deposition within the myocardium such as amyloid fiber or glycolipid, 3) the standardization of two-dimensional echocardiographic interpretation only enabled the quantitative and reliable analysis of tissue texture in cardiomyopathy and various types of myocardial diseases.