Ex Vivo Measurement of Human Blood T1 at 5 T: Dependence on Hematocrit and Oxygenation.

Purpose

To determine human blood longitudinal relaxation time (T1) ex vivo at 5 T using modified Look-Locker inversion recovery (MOLLI), and to evaluate its dependence on hematocrit (Hct) and oxygenation (Y).

Methods

A MOLLI 5(3)3 protocol was simulated across T1 values of 1000-3000 ms to optimize the sampling interval. Aqueous gadolinium phantoms were scanned using optimized MOLLI and multi-inversion-time recovery spin echo (IR-SE) acquisitions to derive a calibration equation. Arm venous blood from 27 healthy volunteers (mean age: 23.7 ± 4.3 years; 10 males) was measured using optimized MOLLI, with three repetitions per sample. Hct and Y were measured by blood gas analysis, and their relationships with calibrated T1 were evaluated.

Results

The minimum root mean square error (RMSE) was obtained at a MOLLI sampling interval of 2000 ms. In phantoms, MOLLI and IR-SE T1 showed a strong linear relationship: T1,IR-SE = 1.185 × T1,MOLLI-93.746 (R2 = 0.998, p < 0.0001). In blood, MOLLI showed high repeatability, with an overall repeatability RMSE of 1.80 ms and a mean coefficient of variation of 0.10% ± 0.08%. Calibrated blood T1 was 1933.6 ± 114.2 ms, range 1727.0-2128.0 ms. Blood 1/T1 dependence on Hct and Y: 1/T1 = 0.447-0.193 × Y + 0.447 × Hct + 0.150 (Y × Hct) (R2 = 0.855, p < 0.0001), yielding an arterial blood T1 of 1971 ms at Y = 0.98 and Hct = 0.42.

Conclusion

Optimized and phantom calibrated MOLLI enabled rapid measurement of human ex vivo blood T1 at 5 T. Blood T1 at 5 T was higher than commonly reported adult values at 3 T and varied with both Hct and Y.