Quantification of left ventricular myocardial mass in humans by nuclear magnetic resonance imaging

Abstract The ability of NMRI to assess LV mass was studied in 20 normal males. By means of a 1.5 Tesla GE superconducting magnet and a standard spinecho pulse sequence, multiple gated short-axis and axial slices of the entire left ventricle were obtained. LV mass was determined by Simpson's rule with the use of a previous experimentally validated method. The weight of the LV apex (subject to partial volume effect in the short-axis images) was derived from axial slices and that of the remaining left ventricle from short-axis slices. The weight of each slice was calculated by multiplying the planimetered surface area of the LV myocardium by slice thickness and by myocardial specific gravity (1.05). Mean ± standard deviation of LV mass and LV mass index were 146 ± 23.1 gm (range 92.3 to 190.4 gm) and 78.4 ± 7.8 gm/m 2 (range 57.7 to 89.4 gm/ m 2 ), respectively. Interobserver agreement as assessed by ICC was high for determining 161 individual slice masses (ICC = 0.99) and for total LV mass (ICC = 0.97). Intraobserver agreement for total LV mass was also high (ICC = 0.96). NMRI-determined LV mass correlated with body surface area: LV mas = 55 + 108 body surface area, r = 0.83; with body weight: LV mass = 26 + 0.77 body weight, r = 0.82; and with body height: LV mass = 262 + 5.9 body height, r = 0.75. Normal limits were developed for these relationships. NMRI-determined LV mass as related to body weight was in agreement with normal limits derived from autopsy literature data. Thus cardiac gated NMR imaging can determine LV mass in humans with a high degree of accuracy. Normal limits of LV mass as determined by NMRI have been developed that will allow precise evaluation of the presence and severity of LV hypertrophy.