Application of the steepest slope model reveals different perfusion territories within the mouse placenta

OBJECTIVES: The steepest slope model is a numerically robust and fast method for perfusion quantification. The purpose of this study was to evaluate if the steepest slope model can be used for quantifying placental perfusion in mice based on dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) datasets.

MATERIAL AND METHODS: T1-weighted DCE MRI was performed in 5 pregnant BALB/c mice on gestation day (gd) 14.5 and in 5 mice on gd 16.5 using a 7T small animal MRI scanner. The placentas were manually delineated in the DCE datasets and the arterial input function (AIF) was selected from the kidney hilus. Placental perfusion was determined on a voxel-by-voxel basis using the steepest slope model. Perfusion was averaged over the entire placenta as well as separately calculated for the high-flow compartment within the central labyrinth zone and for the remaining low-flow placenta tissue. The AIF selection was independently performed by two observers for assessment of inter-observer differences.

RESULTS: Mean perfusion on gd 14.5 was 135 ml/min/100 ml (standard deviation SD: 29 ml/min/100 ml placenta) and 112 ml/min/100 ml on gd 16.5 for the whole placenta (SD: 32 ml/min/100 ml). Perfusion in the high flow compartment in the central labyrinth was significantly higher (p ≤ 0.002) than in the low-flow compartment including the remaining placenta tissue: 184 ml/min/100 ml (SD: 39 ml/min/100 ml) vs. 119 ml/min/100 ml (SD 28 ml/min/100 ml) on gd 14.5 and 158 ml/min/100 ml (SD: 58 ml/min/100 ml) vs. 114 ml/min/100 ml (SD: 52 ml/min/100 ml of placenta) on gd 16.5. The mean relative inter-rater observer difference was 6%.

CONCLUSION: The steepest slope model is a computationally simple method, which allows perfusion quantification in the mouse placenta. Furthermore, the results of this work indicate that the different placental compartments should be analyzed separately to prevent biased results due to averaging.