Back to posts…

CT perfusion (CTP) is a powerful imaging technique that can be utilized in order to distinguish the region of infarction from the ischemic penumbra. CT perfusion algorithms are typically proprietary, and are always dependent on hemodynamic factors. Such factors include presence of extracranial stenosis, presence of sinus rhythm (perfusion may be affected by atrial fibrillation), systemic factors such as hypoperfusion/shock, and more technical factors such as accuracy by which contrast is given, location of selection for both the arterial and venous phase functions, and patient motion during the scan. In an ideal study, CT perfusion is able to demonstrate the 3 typical maps that are used to make decisions, including his cerebral blood flow map [CBF], mean transit time [MTT], and the cerebral blood volume [CBV].

With thrombectomy now occurring in extended time windows (DEFUSE3 and DAWN trials), the role of CT perfusion has been markedly expanded as CT perfusion, when used appropriately, can provide insight into what brain regions can be salvageable from an ischemia and penumbra perspective.

As usual the devil is in the details.

With regards to DEFUSE3 and DAWN, the following table from a topical review by Dr. Alberts, outlines the criteria used to distinguish ischemic core and mismatch volume and the presence of a large vessel occlusion. Similar criteria can be used an MRI imaging (perfusion imaging), however this article will focus on CTP.

Of note, median ischemic core volumes in both trials, are approximately 10 cc, and as evident in the table below, DEFUSE3 included ischemic core volumes up to 70 cc, with a graded approach, dependent on age and NIH stroke scale, as utilized in the DAWN trial.

The ischemic core volume is felt to be estimated by cerebral blood flow less than 30% of normal. This value is felt to correlate to perfusion at the cellular level being less than 20cc/100g/min of brain were critical ischemia occurs. Recall that normal cerebral blood flow is on the order of 50-60cc/100g/min.

Campbell et al., showed that 30% of normal CBF was the optimal threshold to define the ischemic core within a Tmax > 6s (penumbra region), when CT perfusion was processed by standard singular value decomposition. It should be noted that this decomposition mathematically assumes no delay in blood flow from proximal arteries to the ischemic brain tissue whereas T-max and CBF are measured. The threshold of using CBF < 30%, is further supported by data from the SWIFT PRIME trials, where the RAPID software was used, with this threshold of CBF strongly correlating with the final infarct volume at 27 hours inpatient to achieved reperfusion. In the study by Cereda et al., as CBF threshold of less than 38% of normal predicted DWI volumes. However in some patients this threshold significantly overestimated the size of the DWI lesion, and therefore threshold of 30% has been used in the late window EVT trials. Cereda et al., show that a threshold of approximately 38% showed the least mean volume difference (DWI-CTP in cc), of the ischemic core. It should be noted that if a lower percentage is selected such as 25% of normal CBF, that this underestimates the core vs. DWI (i.w. DWI-CTP becomes a larger positive number with lowering of CBF threshold under 38% -> towards 2%).

Similarly T-max greater than 6 seconds has been used to define the ischemic penumbra.

Aside from the fact that these computations are hemodynamic in nature, T-max also provides information with regards to collaterals. And regions of brain with T-max greater than 10 seconds are felt to correlate with regions of poor collateral flow.

The ratio of tissue volume with T-max greater than 10 seconds compared to T-max greater than 6 seconds is referred to as the hypoperfusion intensity ratio, and it is notable that the high rations, greater than 0.5 (i.e. 50%) correlate with poor angiographic collaterals and these patients likely have a March larger ischemic core then visualized, and the score is more likely to rapidly grow.

Therefore in order to use CTP effectively, not only the fact that it is a hemodynamic calculation needs to be considered, the patient has to meet various factors as outlined, but also the details of these thresholds whereby the ischemic core and collateral status are approximated, need to be considered in order to make proper decisions with regards to the information provided by the RAPID software.


  • Albers, Use of Imaging to Select Patients for Late Window Endovascular Therapy,Stroke. 2018;49:00-00. DOI: 10.1161/STROKEAHA.118.021011.
  • Cereda CW, Christensen S, Campbell BC, Mishra NK, Mlynash M, Levi C, et al. A benchmarking tool to evaluate computer tomography perfu- sion infarct core predictions against a DWI standard. J Cereb Blood Flow Metab. 2016;36:1780–1789. doi: 10.1177/0271678X15610586
  • Campbell BC, Christensen S, Levi CR, Desmond PM, Donnan GA, Davis SM, et al. Cerebral blood flow is the optimal CT perfusion param- eter for assessing infarct core. Stroke. 2011;42:3435–3440. doi: 10.1161/ STROKEAHA.111.618355.

Back to posts…