Comparing CT and MRI for the detection of acute stroke
A stroke is a life threatening condition that occurs when the blood supply to part of the brain is cut off. A stroke can affect the way your body works as well as how you think, feel and communicate (NHS, 2017). According to the Stroke Association (2018), there are more than 100,000 strokes in the UK each year. Every 2 seconds, someone worldwide will experience a stroke.
Both CT and MRI are widely available and provide information on the state of the brain, the vessels and brain tissue perfusion in patients. As many institutions can perform either or both imaging techniques, they are not equal and the question often arises as to which one to use in a particular case. Therefore, this blog will explore the two different modalities for imaging acute stroke, based on the review of current literature. Where possible, sensitivity and specificity values will be included for comparison.
| Sensitivity | Specificity | |
| Reference | MRI CT | MRI CT |
| Chalela et al (2007) | 83% 26% | – – |
| Moreau et al (2013) | – 20% | – 98% |
| Mullins et al (2002) | 97% 40% | 100% 92% |
| Lansberg et al (2000) | 57-86% 14-43% | – – |
Chalela et al (2007) carried out a prospective comparison of CT and MRI in a series of patients referred for emergency assessment of suspected acute stroke. Out of 356 patients, 217 who had a final diagnosis of acute stroke were assessed. MRI was found to detect all comparisons of stroke (ischaemic or haemorrhagic) more frequently than CT. Out of the 356 patients, it was shown that MRI detected acute stroke in 164 patients compared to CT, where it was only detected in 35 patients. However, they were both similar in the detection of acute intracranial haemorrhage. From this evidence it can be concluded that MRI (with a sensitivity of 83%) is better than CT (with a sensitivity of 26%) for the detection of acute ischaemia, and can recognise acute and chronic haemorrhage; therefore, it should be the preferred test for accurate diagnosis of patients with suspected acute stroke. For this study, it was stated that the patient sample encompassed a range of diseases that are likely to be encountered in emergency cases for suspected stroke. Therefore, we can say that the results would be applicable to the clinical setting. Despite this, the sample size used was relatively small, meaning that the generalisability of the results are limited. Validity may also be affected due to research being over ten years old.
A more valid and recent study was conducted by Moreau et al (2013) where 347 patients were used to directly compare the yield of acute ischemic lesions on MRI and CT in the emergency diagnosis of suspected minor stroke. Within 24 hours of symptom onset, both scans were compared and it was found that MRI detected acute ischemic lesions in 86% of minor stroke patients and 39% of transient ischemic attack patients, compared to CT which only detected 18% and 8% respectively. Compared to MRI, CT only showed a sensitivity of 20% and a specificity of 98% in identifying acute ischemic lesions. Again, this study only uses a small sample size, however, we can say that the findings are reliable and have high generalisability as they are comparable to the results by Chalela et al (2007) showing that MRI is superior to CT when detecting factors of acute stroke.
Lansberg et al (2000) compared diffusion weighted MRI (DWI) and CT using 19 consecutive stroke patients. Each scan was conducted within 7 hours of stroke symptom onset and was evaluated for acute ischemic lesions like the study above by Moreau et al (2013). It was found that DWI MRI identified acute lesions correctly in all instances, however, on CT it was only identified correctly in 42%-63% of patients. For this study, sensitivity for the detection of acute lesions was higher for MRI (57%-86%) than for CT (14%-43%). Nevertheless, the specificity was equivalent for both. The findings show that again, DWI MRI was more accurate and sensitive for the detection of acute stroke when compared with CT. Although the study is dated, the findings do correspond to the previous studies by Chalela et al (2007) and Moreau et al (2013). Therefore, we can conclude that the results are still valid.
Findings by Mullins et al (2002) closely follow those of Lansberg et al (2000) as it was found that using DW MR images in the early period of a stroke (within 12 hours) resulted in 97% sensitivity and 100% specificity compared with 40% and 92% respectively, with CT. However, it was found that after the 12 hour period, the accuracy of both imaging techniques were equivalent.
Despite the evidence that MRI is the best imaging modality, it can still be said that CT is widely available with a fast acquisition time when compared to MRI (Six, 2018). Additionally, it does remain the golden standard choice for detecting intracranial haemorrhage in the initial evaluation of a stroke (Muir and Santosh, 2005). This can correspond to literature by Barber et al (2005) who suggest that the differences between CT and DWI MR imaging are extremely small. One hundred consecutive stroke patients were used who had undergone both CT and MRI scans within 7 hours of stroke onset. Conclusions from this study suggest that CT is a much faster and more accessible method than MRI, and therefore, is the better neuroimaging modality for the treatment of acute stroke. This study not comparable to previous studies in this blog. However, the level of validity should be considered as there are not enough studies to support its findings. Additionally, only a small sample size was used which reduces how far the results can be generalised. It was also conducted over ten years ago, meaning the findings are outdated.
From the literature, we can conclude that MRI is superior to CT for the imaging of acute stroke. Most sensitivity and specificity values for MRI were seen to be much higher than those for CT. Studies do suggest that using DWI MR imaging is a much better technique for identifying acute ischemic lesions on the brain compared to conventional MR techniques. Despite this, CT is still the primary imaging technique for detecting haemorrhages and the scanning time is much quicker in comparison to MRI. I do believe that more recent research needs to be carried out into which modality is the best for the imaging of acute stroke patients.
References
Barber, P., Hill, M., Eliasziw, M, Demchuk, A., Pexman, J., Hudon, M., Tomanek, A., Frayne, R and Buchan, A. (2005) Imaging of the brain in acute ischaemic stroke: comparison of computed tomography and magnetic resonance diffusion-weighted imaging. Journal of Neurology, Neurosurgery and Psychiatry [online]. 76 (11), pp. 1528-1533. [Accessed 21 November 2018].
Chalela, J.A., Kidwell, C.S., Nentwich, LM, Luby, M., Butman, J.A. and Demchuk, A., et al. (2007) Magnetic Resonance Imaging and Computed Tomography in Emergency Assessment of Patients with Suspected Acute Stroke: A Prospective Comparison. The Lancet [online]. 369 (9558), pp. 293-298. [Accessed 20 November 2018].
Lansberg, F.G., Albers, G.W., Beaulieu, C., Marks, M.P. (2000) Comparison of diffusion-weighted MRI and CT in acute stroke. Neurology [online]. 54 (8) [Accessed 20 November 2018].
Moreau, F., Asdaghi, N., Modi, J., Goyal, M. and Coutts, S.B. (2013) Magnetic Resonance Imaging Versus Computed Tomography in Transient Ischemic Attack and Minor Stroke: The More Υou See the More You Know. Cerebrovascular Diseases Extra [online]. 3 (1), pp. 130-136. [Accessed 20 November 2018].
Muir, K.W. and Santosh, C. (2005) Imaging of Acute Stroke and Transient Ischaemic Attack. Journal of Neurology, Neurosurgery and Psychiatry [online]. 76 (3), p. 19. [Accessed 20 November 2018].
Mullins, ME, Schaefer, PW, Sorensen, A.G., Halpern, E.F., Ay, H., He, J., J, W., Koroshetz, R and Gonzalez, G. (2002) Ct and Conventional and Diffusion-weighted Mr Imaging in Acute Stroke: Study in 691 Patients at Presentation to the Emergency Department. Radiology [online]. 224 (2), pp. 10-11. [Accessed 21 November 2018].
NHS (2017) Overview Stroke. Available from: https://www.nhs.uk/conditions/stroke/ [Accessed 20 November 2018].
Six, 0. (2018) The role of CT and MR in stroke patients. Quantib [blog]. 14 March. Available from: https://www.quantib.com/blog/the-role-of-ct-and-mr-in-stroke-patients [Accessed 21 November 2018].
Stroke Association (2018) Stroke statistics. Available from: https://www.stroke.org.uk/system/files/sotn_2018.pdf [Accessed 20 November 2018].
Serena's Blog 
Hi Serena.
Another good blog. You’ve definitely got your head around them! You’ve done well to emphasise the issue of the validity of your older references, the next step is to point out why. Sequences in MRI are constantly evolving which improves the sensitivity and specificity of its clinical use. Following on from your conclusions, current information is out there with regards to acute stroke imaging. Brain Imaging in Acute Ischemic Stroke—MRI or CT? by Audebert & Fiebach considers different CT and MI techniques. You have used current sources at points to back up some of the older references you have used which was a good touch and shows you are aware of the issues the older sources can bring to your work.
The National Stroke Strategy and NHS England have highlighted this fact and as such noted that the use of MRI and CT in diagnosis should be reviewed regularly. This is also the reason why you do not find preferred imaging modality in the current NICE guidelines. https://www.strokeaudit.org/SupportFiles/Documents/Guidelines/2016-National-Clinical-Guideline-for-Stroke-5t-(1).aspx – this document highlights at point 3.4 on page 39 that imaging needs to be performed within an hour of the patient arriving at hospital. how would that affect which modality you use? would it be the same across all hospitals?
For your next blog, try and keep your references current to avoid falling into the trap of constantly having to say your sources may not be valid due to their age. By all means, you can use older sources to make a point, e.g. to show something hasn’t changed by using a current reference with it or if it a seminal piece of information but try and stay current.
Cornelius
LikeLike