Introduction: It has been observed that there are limitations to the detection of subdural blood by various available MRI sequences even blood that is macroscopically evident at surgery. This experiment was therefore carried out to determine the threshold concentration that the MRI maybe able to detect blood.
Methods: Expired blood from the blood bank was obtained and the hematocrit was determined in the standard fashion. The blood was then serially diluted and the again the hematocrit determined again in each of the serially diluted specimens and they were placed in 1-inch wide test tubes and scanned in a 3 tesla MRI. The images were then reviewed and the signal intensity in each sample was measured 3 times using the reason of interest function and the average calculated for each sample. The MRI technologist and radiologist were blinded to the concentrations of blood in the test tubes.
Results: Visual inspection showed that confident determination of the presence of blood could only be made at concentrations over 40% hematocrit. However, there was a strong correlation of the concentration of red cells in each sample with the signal intensity with a positive correlation coefficient (r2) of 0.99 in the T1W images and a negative correlation coefficient of 0.99 in the ADC sequence
Conclusions: The findings confirm the clinical observation that MRI lacks sensitivity in detecting blood especially in the subdural space. Further studies are needed to quantify the sensitivity of MRI when CSF dilutes subdural blood in vivo.
Patient Care: Cases of subdural effusions in spontaneous or abusive trauma especially in children need to be interpreted with caution noting the lack of sensitivity of MRI in detecting blood
Learning Objectives: Highlight the limitation of MRI in detecting blood in the subdural space. Show the correlation of signal intensity and blood concentration highlighting that there is no clear cut-off point.