Quantitative CBF maps derived from pseudo-continuous ASL (pCASL) may be useful in assessing stroke risk in patients afflicted with sickle cell anemia (SCA). The computation of CBF¹ requires knowledge of the T1 relaxation time of blood which has been elegantly characterized for normal subjects as a function of oxygenation level, hematocrit, and field strength^2,3. However, longitudinal relaxation times have not been characterized in SCA patients whose hemoglobin consists of the variant hemoglobin S (HbS) in addition to normal hemoglobin A (HbA). In this study, we measure the T1 relaxation times of SCA blood over a normal range of hematocrit and percent HbS.Experiment. Six samples (n=3 SCA; n=3 non-SCA) of human venous blood were collected into heparinized tubes from volunteers who provided informed, written consent. For SCA blood, laboratory tests were performed to obtain the hematocrit and percent HbS values. For normal blood, hematocrit values were obtained using an i-STAT handheld blood analyzer (Abbott, Ottawa, Canada). Blood oxygenation measurements were obtained for all six samples using the blood analyzer. MR measurements were performed on the blood samples at 3T (Philips), and care was taken to ensure in vivo temperature conditions. Prior to scanning, the temperature of the blood samples was stabilized to approximately 37 °C using a heated water bath. During scanning, the tubes were positioned inside a jar of peanut butter which was also previously heated to 37 °C. The container was then wrapped in a warm water circulation system and placed in the scanner. T₁-weighted MR images were acquired using an inversion recovery sequence: TE=90ms; TI=30ms, 60ms,100ms, 250ms, 500ms, 1000ms, 2000ms, 3000ms, 5000ms with constant recovery time=8000ms; in-plane spatial resolution=1.25x1.25mm². Analysis. The T₁ of each sample was computed by fitting the mean magnitude MR signal from a region inside the blood sample at each TI using a two-parameter model and a non-linear least squares algorithm⁴.For the normal blood samples, the mean (±standard deviation) values were 39% (±0.6) for the hematocrit and 1.55s (±0.11) for the T₁. For the SCA blood samples, the mean (±standard deviation) values were 32% (±7) for the hematocrit, 78.8% (±4.8) for the percent HbS, and 1.47s (±0.13) for the T₁. Representative T₁ maps (Figure 1A) and the two-parameter fitting for extracting T₁ (Figure 1B) are shown for representative normal and SCA blood samples.CBF values computed from quantitative pCASL measurements may be useful in assessing stroke risk in SCA patients. However, T₁ relaxation of blood from SCA patients, which contains HbS in addition to HbA, must be characterized before CBF can be reliably quantified. For similar hematocrit, oxygenation, and temperature, T₁ relaxation times of SCA blood appear similar to those of normal blood, and assumption of normal blood T₁ for SCA patients would result in a relatively low error in CBF of approximately 8.6%. These results should be considered in the context of the following limitations. First, the temperature of the samples during scanning was lower than 37 °C but was identical for both normal and SCA blood. Next, only venous blood oxygenation was considered. Finally, the experiments were conducted on ex vivo samples; in vivo SCA blood may have slightly different T1 relaxation characteristics due to flow velocity, turbulence, and other physiological factors. Ongoing work involves examining the dependence of the T1 and T2 times of SCA blood on hematocrit and oxygenation.T₁ relaxation times in SCA blood appear to be similar to those observed in normal blood. Computation of CBF using pCASLfor stroke risk assessment in SCA patients may not be affected by the assumption of normal blood T₁ relaxation.