Subdural haematoma (SDH) resulting from traumatic brain injury relating to non accidental head injury (NAHI) is unfortunately relatively common in the UK at 36 per 100000 incidence in children < 6 months old¹. SDH as a result of NAHI occurs more frequently in younger babies partly due to larger subdural space and more susceptibility to subdural bleeds. Currently SDH are classified as acute (< 1 week), subacute (> 1 week) or chronic (> 1 month) with further sub-categories reported². However, variations between radiologists from different hospitals in dating SDH have also been observed³. Fetal blood has a higher affinity to oxygen compared to adult blood to allow gas exchange between maternal and fetal circulations within the placenta, with fetal blood being dominant up until 6 months of age⁴. The aim of this study was to determine if MRI could detect any difference between fetal and adult blood for relaxation time constants T­₁ and T₂^* in simulated SDH over a 28 day period.

16 consenting full-term pregnant women allowed collection of 10ml of fetal blood extracted from the umbilical cord from the placenta during arranged caesarean section. An adult blood sample was collected from a gender matched volunteer within 30 minutes of each fetal blood sample being collected. 10ml of blood was placed inside a 15cm semi-permeable membrane to simulate a blood clot seen in the brain. The filled membrane was suspended inside a 50ml sample tube containing 45ml of artificial cerebral spinal fluid (aCSF), prepared using the method described by Alzet⁵. Filled sample tubes were maintained at 37°C in a water bath for the study duration.

Data acquisition: Data were acquired on a 1.5 T Avanto system (Siemens, Erlangen, Germany) using a coronal plane and 16 channel body flex coil. A mGE sequence with parameters: TR:1000ms; 10 TEs between 12 and 62ms with 6ms spacing; FA:25° and bandwidth (BW):260Hz/pixel was used for T₂^* quantification. An IR-TSE sequence with parameters TR/TE:3882/12ms was used to acquire 5 datasets with IR:150, 250, 400, 750, 1000ms; FA:180° and BW:130Hz/pixel, for T₁ quantification. Both sequences had 300mm FOV and resolution: 1.17 mm²; 10 slices at 4mm thickness and 0.4mm slice gap. Data were acquired once daily for 28 days.

Data Processing: Regions of interest (ROI) were drawn using a PACS workstation (IMPAX 5, Agfa, Belgium) within the resulting blood clot in each tube and data fitted for T₁ and T₂^* time constants using least squares fitting in Matlab (MathWorks, UK). Data were averaged into 7 day bins for further statistical analysis using paired T- and Wilcoxon signed ranks test using Microsoft Excel.

Calculated T_­2^* values (figure 1) showed significant differences in week 1 (p<0.0001) between fetal and adult blood. Calculated T₁ values (figure 2) showed significant differences (p<0.01) in weeks 1,3 and 4 between fetal and adult blood. The relationship between fetal and adult blood T₂^* and T₁ can be expressed as y = 0.85x + 8.70ms and y = 1.03 - 63.58ms respectively. Comparing fetal and adult blood T₂^* values for all time points showed a significant difference (p<0.00001).Numerous studies have already examined the differences in appearance of SDH in adults using MRI^6,7, but this is believed to be the first study to compare differences between fetal and adult blood, albeit in an artificial model. The aim of this study was to quantify relaxation time constants T₁ and T₂^* for fetal and adult blood over a period of 28 days with significant differences found between the two blood groups. It was expected that as the blood samples became desaturated, blood plasma permeated into the aCSF and ferretin and hemosiderin concentrations increased within the blood sample, T₁ and T₂^* values would decrease over time. However, this was not the case with increases in both relaxation time constant values seen in fetal and adult blood. The increase in values could be due to degradation of deoxyhaemoglobin to methaemoglobin. These results have particular importance (with potential medicolegal implications) in cases of NAHI in infants where SDH is a primary clinical presentation and adult models are used to age the SDH.