Facebook Pixel Code
Banner image
Fetal MRI


Figure of fetal blood vessels
T-2 map of the fetal blood vessels.

Fetal magnetic resonance oximetry has the potential to improve upon current, ultrasound-based methods for the diagnosis of fetal hypoxia. Unlike ultrasound, which detects hypoxia by the measurement of adaptive blood flow changes (i.e. brain sparing), MRI is directly sensitive to blood oxygen content. This phenomenon is exploited in a variety of of MRI techniques, which are well established for postnatal imaging, such as Blood Oxygenation Level Dependent (BOLD) MRI and vascular relaxometry/susceptometry.

Correct estimation of blood oxygen content from relaxometry/susceptibility data requires an accurate biophysical model and calibration, which provides the relationship between MRI properties (i.e. T1, T2, and susceptibility, ) and blood properties (hematocrit, Hct, and oxygen saturation, sO2). While such relationships are established for adult blood, data on fetal blood properties is sparse.  

Fetal blood differs from adult blood in several respects, most notably in the structure of hemoglobin, the oxygen transport protein.  Our recent work has therefore focused on the comprehensive characterization of the MRI properties of fetal blood at 1.5T and 3T field strengths [1-3]. Python code for estimation of hematocrit and oxygen saturation from measurements of blood vessel T1 and T2 relaxation times is available for download at: https://github.com/shportnoy/oximetry_calculator.

Ongoing research in our group is dedicated to improving the robustness of fetal blood vessel T1 and T2 measurements (which can then be used to estimate Hct and sO2). Accurate measurement of relaxation times in fetal vessels is uniquely challenging, requiring sufficient resolution, as well as robustness to fetal motion and pulsatile blood flow, all in the absence of a cardiac gating signal with which to synchronize the acquisition. Future innovations in this area will leverage the speed and flexibility offered by non-Cartesian acquisitions and iterative reconstruction strategies.

[1] Portnoy S, Osmond M, Zhu MY, Seed M, Sled JG, Macgowan CK. Relaxation properties of human umbilical cord blood at 1.5 Tesla. Magn Reson Med. 77(4):1678-1690 (2017) – doi: 10.1002/mrm.26231

[2] Portnoy S, Seed M, Sled JG and Macgowan CK. Non-invasive evaluation of blood oxygen saturation and hematocrit from T1 and T2 relaxation times: In-vitro validation in fetal blood. Magn Reson Med. 78(6):2352-2359 (2017) – doi:10.1002/mrm.26599

[3] Portnoy S, Milligan N, Seed M, Sled JG, Macgowan CK. Human umbilical cord blood relaxation times and susceptibility at 3 T. Magn Reson Med. 79:3194-3206 (2018) – doi: 10.1002/mrm.26978

[4] Sun L, Macgowan CK, Sled JG, Manlhiot C, Porayette P, Grosse-Wortmann L, Jaeggi E, McCrindle BW, Kingdom J, Hickey E, Miller S, Seed M. Reduced fetal cerebral oxygen consumption Is associated with smaller brain size in fetuses with congenital heart disease. Circulation 131:1313-1323 (2015) – doi:10.1161/CIRCULATIONAHA.114.013051

[5] Zhu MY, Milligan N, Keating S, Windrim R, Keunen J, Thakur V, Ohman A, Portnoy S, Sled JG, Kelly E, Yoo SJ, Gross-Wortmann L, Jaeggi E, Macgowan CK, Kingdom JC, Seed M. The hemodynamics of late onset intrauterine growth restriction by MRI. Am J Obstet Gynecol. 214(3):367.e1-367.e17 (2015) – doi:10.1016/j.ajog.2015.10.004