Brain perfusion imaging in patients with status epilepticus, seizures, and IIC patterns Authors: Khalil S. Husari, Michael Porambo, Vivek S. Yedavalli, Peter W. Kaplan Published: 2026 ABSTRACT Highlights•Hyperperfusion on brain perfusion imaging is specific for seizure conditions.•Their sensitivity for SE is enhanced when thalamic hyperperfusion is present.•Brain perfusion imaging accuracy is variable in diagnosing seizures. Introduction Changes in cerebral perfusion during both the ictal and interictal phases of an epileptic seizure are well-established phenomena. Penfield noted an increase in cerebral regional perfusion along with focal changes in vasomotor dynamics during epileptic seizures. (Penfield, 1933, Leblanc, 2019) Seizures trigger a sharp increase in cortical neuronal activity, metabolism, and oxygen consumption. (Guerriero and Gaillard, 2019) This heightened demand is typically met by an increase in regional cerebral blood flow (CBF), driven by neurovascular coupling mechanisms. Hence, cellular energy status is initially maintained. However, in prolonged seizures and in other certain instances, there is uncoupling of glucose metabolism and blood flow resulting in regional tissue hypoxia and anaerobic glycolysis. (Guerriero and Gaillard, 2019) Brain perfusion imaging (computed tomography perfusion (CTP) and Magnetic Resonance Perfusion (MRP) can be used to depict the changes in regional cerebral blood flow associated with seizures and status epilepticus. (Guerriero and Gaillard, 2019). In the last ten years, the application of CTP and MRP have grown significantly in evaluating patients with acute stroke and other neurological emergencies. (Menon, 2020, Weinstein et al., 2022) As these imaging modalities became more widely available and accessible, their use has been extended to investigate stroke mimics, including conditions like seizures and status epilepticus. (Ahmed et al., 2023, Khalili et al., 2023) Multiple studies have investigated the sensitivity and specificity of brain perfusion imaging in different epileptic conditions (seizures, SE, and ictal-interictal continuum patterns), with variable findings. (Merli et al., 2023, Ameen Ahmad et al., 2024, Merli et al., 2024) Some clinicians have used functional neuroimaging results in treatment algorithms and to decide whether to escalate or de-escalate antiseizure therapy in patients with status epilepticus and ictal-interictal continuum patterns. (Claassen, 2009, Husari and Johnson, 2020) Despite these advances, a recent and comprehensive review that synthesizes the latest data on this topic remains lacking. In this review article, we summarize the current studies examining the role of CTP and MRP in patients with seizures and status epilepticus. Our review will be outlined based on the epileptic condition of 1) status epilepticus, 2) seizures, and 3) electrographic seizures and IIC patterns. Basics of CTP and MRP CTP images are acquired after the injection of IV contrast. Images are obtained every 1–3 s, before, during and after the injection of intravenous contrast. Sequential images are captured to track contrast passage through the cerebral vasculature, and then summated to compute perfusion maps. Physiologically, brain tissue with intact autoregulation and perfusion demonstrates rapid contrast arrival and washout, whereas ischemic or hypoperfused tissue demonstrates delayed arrival, delayed washout, and reduced flow. The most clinically utilized perfusion maps are cerebral blood flow (CBF, ml/100 g/minute), cerebral blood volume (CBV, ml/100 g), mean transit time (MTT; seconds), and time to maximum of the residual function (Tmax; seconds). (Menon, 2020);(Khalili, Wang et al., 2023). The most common MRP methods in neuroimaging are arterial spin labeling (ASL) and dynamic susceptibility contrast (DSC). (Daftari Besheli, Ahmed et al., 2022) ASL is a non-contrast method and utilizes magnetically labeled inflowing arterial blood water as an endogenous tracer. A radiofrequency (RF) pulse inverts or “labels” the longitudinal magnetization of inflowing arterial blood at the level of the carotid and vertebral arteries in the neck. After a short post-labeling delay, images of the brain are acquired. Control images are obtained without labeling the inflowing blood, and the labeled signal is subtracted from the control images to yield a perfusion-weighted signal, proportional to the amount of labeled blood delivered to tissue. (Daftari Besheli, Ahmed et al., 2022) As a non-contrast technique, ASL has utility in its repeatability (especially in epilepsy patients) as well as in pediatric and renally impaired patients and those with contrast allergies. DSC is a contrast-enhanced method which involves tracking the signal drop on T2*-weighted images during the passage of a gadolinium bolus. The susceptibility-induced signal loss corresponds to contrast concentration, allowing calculation of CBF, CBV, and MTT. During the ictal phase, there is a relative increase in the cerebral perfusion due to an increase in cerebral metabolism and neurovascular coupling. This regional hyperperfusion is typically characterized by increased relative cerebral blood flow (rCBF) and relative cerebral blood volume (rCBV) and decreased measures of delay (MTT and Tmax). (Khalili, Wang et al., 2023) In juxtaposition to acute infarct, these findings are primarily cortical in a non-territorial distribution without corresponding vascular abnormality on angiographic studies. During the interictal phase, these areas typically demonstrate a decrease in CBF and CBV and an increase in MTT. Table 1 summarizes the perfusion findings in epileptic conditions in comparison to acute stroke.Table 1Perfusion findings in various epileptic conditions, adapted from (Austein, Huhndorf et al., 2018).Ictal stateInterictalAcute stroke – infarct coreAcute stroke − ischemic penumbraCTAHyperemia and engorged vesselsPresence of vessel occlusionPresence of vessel occlusionCTP/MRP findingsCortical +/- thalamic hyperperfusion is commonHypoperfusion (cortical & subcortical) or normo-perfusionHypoperfusionHypoperfusionAbnormality restricted to vascular territoryNoNoYesYesrCBF↑↓or →↓↓↓rCBV↑↓ or →↓↓ (core)↑ or →MTT↓(faster transit)↑↑↑↑TTP↓(faster transit)↑↑↑↑Tmax↓↑↑↑↑CTA: CT angiography, CTP: CT Perfusion, MRP: MR perfusion, rCBF: regional cerebral blood flow, rCBV: regional cerebral blood volume, MTT: mean transit time, TTP: time to peak, Tmax: Time-to-Maximum. Table 1Perfusion findings in various epileptic conditions, adapted from (Austein, Huhndorf et al., 2018).Ictal stateInterictalAcute stroke – infarct coreAcute stroke − ischemic penumbraCTAHyperemia and engorged vesselsPresence of vessel occlusionPresence of vessel occlusionCTP/MRP findingsCortical +/- thalamic hyperperfusion is commonHypoperfusion (cortical & subcortical) or normo-perfusionHypoperfusionHypoperfusionAbnormality restricted to vascular territoryNoNoYesYesrCBF↑↓or →↓↓↓rCBV↑↓ or →↓↓ (core)↑ or →MTT↓(faster transit)↑↑↑↑TTP↓(faster transit)↑↑↑↑Tmax↓↑↑↑↑CTA: CT angiography, CTP: CT Perfusion, MRP: MR perfusion, rCBF: regional cerebral blood flow, rCBV: regional cerebral blood volume, MTT: mean transit time, TTP: time to peak, Tmax: Time-to-Maximum. Perfusion findings in various epileptic conditions, adapted from (Austein, Huhndorf et al., 2018). CTA: CT angiography, CTP: CT Perfusion, MRP: MR perfusion, rCBF: regional cerebral blood flow, rCBV: regional cerebral blood volume, MTT: mean transit time, TTP: time to peak, Tmax: Time-to-Maximum. Methods This narrative literature review was conducted using PubMed database for articles that included the following combination keywords in their title and abstract.: (“CT perfusion” OR ”computed tomography perfusion“) AND (”seizure“ OR ”seizures “OR “status epilepticus” OR SE OR “ictal-interictal continuum “OR IIC) OR (”MR perfusion OR “magnetic resonance perfusion” OR ”perfusion MRI“) AND (”seizure“ OR ”seizures“ OR ”status epilepticus“ OR SE OR ”ictal-interictal continuum“ OR IIC). Original articles and reviews were included. Article and review references were screened for additional sources. Only articles and reviews published in English were used in this review. CT perfusion in status epilepticus There have been several studies that explored the diagnostic value of CTP in patients with SE, particularly in emergency settings. Most of these studies show high specificity for focal hyperperfusion with variable sensitivity. Across the board, studies have shown good sensitivity and specificity for identifying patients in status epilepticus. A summary of the studies evaluating CTP in status epilepticus is presented in Table 2.Table 2Summary of studies evaluating CTP in status epilepticus.ReferenceTypeSynopsisTimingFindings(Hauf, Slotboom et al., 2009)Case series19 patients with seizures/AMS;CTP/EEG 2 h78% of NCSE had hyperperfusion; 30% postictal had hypoperfusion with Todd's paresis, 70% had normal CTP(Strambo, Rey et al., 2018)Case series43 epileptic stroke mimicsCTP/EEG 24 h49% had abnormal CTP (30% hyper, 19% hypo).30% (13 patients) had SE.69% of SE patients had hyperperfusion4 with seizures (not SE) had hyperperfusion(Gonzalez-Cuevas, Coscojuela et al., 2019)Case-control19 SE vs 10 controls (most w/ single seizures)CTP/EEG mean ≈ 40 min in SE group79% of SE had hyperperfusion; 90% specificity, 79% sensitivity for SEIn control group, 60% had normal perfusion, 30% hypoperfusion, 10% had hyperperfusion(Khoo, Kim et al., 2020)Case control27 seizure/SE vs 20 S.Symptom onset/CTP mean 5 h63% had hyperperfusion, 7% hypoperfusion74% multi-territorial, 58% thalamic involvement(Giovannini, Malagoli et al., 2021)Case series21 NCSE patientsSymptom onset/CTP median 3 h86% had hyperperfusion; 14% had normal perfusion.Presence of thalamic & cortical hyperperfusion is highly diagnostic of NCSE.(Romoli, Merli et al., 2022)Case control17 NCSE vs matched cohort of stroke patientsSymptom onset/CTP median 2 hCTP/EEG median 30 min65% of NCSE had hyperperfusion vs 0% stroke: 65% sensitivity and 100% specificity(Restrepo-Vera, Coscojuela et al., 2022)Case series27 seizure and 20 SE patients; CTP at 3 h medianSymptom onset/CTP within 3 h38% (18 patients) had hyperperfusion (12 had SE, 6 single seizure)60% of SE had hyperperfusion compared to 22% of single seizure patients.Thalamic hyperperfusion was present only in SE patients.Hyperperfusion had a sensitivity of 60% and 77.8% specificity for the detection of SE.(Merli, Galluzzo et al., 2022)Case reportPost stroke epilepsy with acute onset aphasia.n/aSerial CTP/EEG showed ictal hyperperfusion and NCSE on EEG, pre/postictal hypoperfusion with EEG slowing(Merli, Romoli et al., 2023)Case control21 NCSE vs 56 controlsSymptom/CTP, median 9 hCTP and EEG within 2 h of presentation and each other95% of NCSE had hyperperfusion; AUROC 0.79, high NPV 0.97(Lee, Sohn et al., 2024)Case series73 SE patients; multimodal imagingSymptom/CTP, median 5.8 h26% had CTP hyperperfusion(Merli, Romoli et al., 2025)Case series15 post-stroke SESymptom/CTP, median 3 h100% hyperperfusion.Most had had extra-lesional hyperperfusion (either perilesional to extra-lesional)5 had thalamic hyperperfusion, 5 had hyperperfusion in multiple areas.In 9/15, EEG showed focal NCSE per Salzburg criteriaAMS: altered mental status, NCSE: non-convulsive status epilepticus, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, AUROC: Area Under the Receiver Operating Characteristic curve, NPV: Negative Predictive Value. Table 2Summary of studies evaluating CTP in status epilepticus.ReferenceTypeSynopsisTimingFindings(Hauf, Slotboom et al., 2009)Case series19 patients with seizures/AMS;CTP/EEG 2 h78% of NCSE had hyperperfusion; 30% postictal had hypoperfusion with Todd's paresis, 70% had normal CTP(Strambo, Rey et al., 2018)Case series43 epileptic stroke mimicsCTP/EEG 24 h49% had abnormal CTP (30% hyper, 19% hypo).30% (13 patients) had SE.69% of SE patients had hyperperfusion4 with seizures (not SE) had hyperperfusion(Gonzalez-Cuevas, Coscojuela et al., 2019)Case-control19 SE vs 10 controls (most w/ single seizures)CTP/EEG mean ≈ 40 min in SE group79% of SE had hyperperfusion; 90% specificity, 79% sensitivity for SEIn control group, 60% had normal perfusion, 30% hypoperfusion, 10% had hyperperfusion(Khoo, Kim et al., 2020)Case control27 seizure/SE vs 20 S.Symptom onset/CTP mean 5 h63% had hyperperfusion, 7% hypoperfusion74% multi-territorial, 58% thalamic involvement(Giovannini, Malagoli et al., 2021)Case series21 NCSE patientsSymptom onset/CTP median 3 h86% had hyperperfusion; 14% had normal perfusion.Presence of thalamic & cortical hyperperfusion is highly diagnostic of NCSE.(Romoli, Merli et al., 2022)Case control17 NCSE vs matched cohort of stroke patientsSymptom onset/CTP median 2 hCTP/EEG median 30 min65% of NCSE had hyperperfusion vs 0% stroke: 65% sensitivity and 100% specificity(Restrepo-Vera, Coscojuela et al., 2022)Case series27 seizure and 20 SE patients; CTP at 3 h medianSymptom onset/CTP within 3 h38% (18 patients) had hyperperfusion (12 had SE, 6 single seizure)60% of SE had hyperperfusion compared to 22% of single seizure patients.Thalamic hyperperfusion was present only in SE patients.Hyperperfusion had a sensitivity of 60% and 77.8% specificity for the detection of SE.(Merli, Galluzzo et al., 2022)Case reportPost stroke epilepsy with acute onset aphasia.n/aSerial CTP/EEG showed ictal hyperperfusion and NCSE on EEG, pre/postictal hypoperfusion with EEG slowing(Merli, Romoli et al., 2023)Case control21 NCSE vs 56 controlsSymptom/CTP, median 9 hCTP and EEG within 2 h of presentation and each other95% of NCSE had hyperperfusion; AUROC 0.79, high NPV 0.97(Lee, Sohn et al., 2024)Case series73 SE patients; multimodal imagingSymptom/CTP, median 5.8 h26% had CTP hyperperfusion(Merli, Romoli et al., 2025)Case series15 post-stroke SESymptom/CTP, median 3 h100% hyperperfusion.Most had had extra-lesional hyperperfusion (either perilesional to extra-lesional)5 had thalamic hyperperfusion, 5 had hyperperfusion in multiple areas.In 9/15, EEG showed focal NCSE per Salzburg criteriaAMS: altered mental status, NCSE: non-convulsive status epilepticus, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, AUROC: Area Under the Receiver Operating Characteristic curve, NPV: Negative Predictive Value. Summary of studies evaluating CTP in status epilepticus. AMS: altered mental status, NCSE: non-convulsive status epilepticus, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, AUROC: Area Under the Receiver Operating Characteristic curve, NPV: Negative Predictive Value. One of the earliest studies was conducted by Hauf et al., in which they evaluated nine patients with non-convulsive status epilepticus (NCSE), comparing them to ten patients in postictal state. CTP regional hyperperfusion was present in 7/9 NCSE patients (78%), while none of the postictal patients had hyperperfusion. Patients with NCSE had significantly increased rCBF, increased rCBV, and decreased MTT. (Hauf, Slotboom et al., 2009) Of the ten postictal patients, three had regional hypoperfusion corresponding to Todd’s paresis, while the remaining had normal perfusion. (Hauf, Slotboom et al., 2009). In a study by Merli et al., 21 patients with NCSE were compared to a control group of 57 patients. The control group included patients with seizures, postictal states, encephalopathy, and migraine aura. CTP hyperperfusion was present in 95.2% of NCSE patients compared to 37.5% of non-NCSE patients (20/21 vs. 21/56, p < 0.001). Overall, CTP hyperperfusion demonstrated a 79% diagnostic accuracy for NCSE, with 95% specificity and a 97% negative predictive value (NPV). Most patients with CTP hyperperfusion but without a NCSE diagnosis were found to have either seizures or were in a postictal state, raising the possibility that some of these cases may have represented undiagnosed NCSE. (Merli, Romoli et al., 2023). Giovannini et al. evaluated 21 patients with NCSE with CTP and EEG obtained within a few hours of the CTP. Hyperperfusion was present in 18 patients (86%); 17 patients had cortical hyperperfusion, and eight patients had ipsilateral thalamic hyperperfusion. Three patients had a normal perfusion pattern. Among the 18 patients presenting with hyperperfusion, 16 (89%) exhibited an ictal EEG pattern, while the remaining two patients showed postictal focal slowing with or without interictal epileptiform discharges (status post benzodiazepine treatment prior to EEG acquisition). Among the three patients with normal perfusion, 2 had ictal EEG activity and met criteria for definite NCSE. (Giovannini, Malagoli et al., 2021). Romoli et al. compared 17 NCSE patients with a 1:1 matched group of patients with acute ischemic stroke. The authors assessed CTP using both conventional analysis and a novel reverse Tmax (rTmax) analysis. Hyperperfusion was found in 64.7% of NCSE patients and in none of the stroke patients, yielding 100% specificity. (Romoli, Merli et al., 2022). CT perfusion in seizures The findings of CTP in patients with seizures have been more heterogenous across various studies. A major contributing factor is likely the difference in timing between onset of symptoms and acquisition of CTP. The switch in regional cerebral perfusion from hyperperfusion to hypoperfusion is a dynamic process and varies in timing and in specific brain regions. (Newton et al., 1992, Duncan et al., 1993, Gaxiola-Valdez et al., 2017) Based on SPECT studies in temporal lobe epilepsy, the perfusion switch typically occurs within 2 min, but a lagging hyperperfusion can persist for several minutes. (Newton et al., 1992, Duncan et al., 1993, Leonhardt et al., 2005, Gaxiola-Valdez et al., 2017) For example, in patients with focal impaired aware seizures scanned within 12 min of seizures, there was a decrease in the perfusion in the hippocampus compared to lagging hyperperfusion in the parahippocampus. (Leonhardt, de Greiff et al., 2005) A summary of the studies evaluating CTP in seizures is presented in Table 3.Table 3Summary of studies evaluating CTP in seizures.ReferenceTypeSynopsisTimingFindings(Wiest, von Bredow et al., 2006)Case series15 patientsCTP/EEG 2 hVariable perfusion patterns: 3 subtle SE had hyperperfusion (2 EEG concordant); 12 postictal: 6 hypoperfusion with EEG focal slowing, 6 normal (with normal EEG)(Gelfand, Wintermark et al., 2010)Case series27 seizure patientsCTP/presentation 72 h37% had abnormal CTP; only 4% had hyperperfusion.Imaging within 2 h after seizure termination was associated with a higher likelihood of abnormal CTP.(Payabvash, Oswood et al., 2015)Case series18 seizure patientsSymptom/CTP 3 h14/18 patients (78%) had unilateral hyperperfusion. None had hypoperfusion(Van Cauwenberge, Dekeyzer et al., 2018)Case series133 seizure patients: 50% Todd’s phenomenon 28% postictal w/out Todd’s22% SE/recurrent seizuresSymptom/CTP1.9 hPatients with SE/recurrent seizures had higher incidence of hyperperfusion, 38% vs 11–16% (without or w/ Todd’s).CTP hyperperfusion had 38% sensitivity and 86% specificity for detecting ictal activity with an NPV of 83%.(Austein, Huhndorf et al., 2018)Case series37 with seizure-related stroke mimicsSymptom/CTP within 6 h, ¾ within 3 h60% hyperperfusion, 41% hypoperfusionHyperperfusion in thalamus or mesial temporal structures was significantly correlated with epileptiform EEG activity.On follow up MRI, only patients with CTP hyperperfusion demonstrated seizure-related signal abnormality in the thalamus or hippocampus.(Manganotti, Furlanis et al., 2019)Case series11 epileptic aphasic patientsSymptom/CTP within 4.5 hEEG within 24 h8/11 (73%) had CTP hyperperfusion and sharp waves on EEG.3/11 (27%) had hypoperfusion and slowing on EEG.(Lucas, Gariel et al., 2021)Case control45 seizure vs 50 ischemic strokeSymptom/CTP within 4.5 hEEG during hospitalizationHyperperfusion was more frequent in the seizure group compared to stroke group (36% vs 2%).Hyperperfusion had high specificity of 98% but weak sensitivity of 35% for diagnosing seizures.Specificity and PPV improved if the perfusion pattern was not restricted to a vascular territory(Gonzalez-Martinez, Trillo et al., 2021)Case series94 seizure and 21 SE, out of 284 S mimicsSymptom/CTP, mean 2.9 h50% of patients with seizures or SE had abnormal CTP.CBV was increased in 17% of seizure patients and 19% of SE.(Prodi, Danieli et al., 2022)Case series31 seizure patients out of 401 S/stroke mimicsn/aAbnormal CTP was present in 42% of seizure patients.23% hyperperfusion, 19% hypoperfusion.(Ameen Ahmad, Primiani et al., 2024)Case series103 patientsCTP/EEG within 6 h35 patients had clinical seizures: 1/3 had hyperperfusion, 1/3 hypoperfusion, 1/3 normal perfusion.CTP hyperperfusion had low sensitivity of 34% but high specificity 96% for identifying patients with clinical seizures.Abnormal CTP had moderate sensitivity of 69% for identifying clinical seizures.(Rapai Kasini, Payus et al., 2025)Case control31 seizures vs 31 LVO strokesSymptom/CTP within 48 hOnly 19% seizure patients had hyperperfusion; most had normal (39%) or hypoperfusion (36%).LVO: large vessel occlusion, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, PPV: Positive Predictive Value, NPV: Negative Predictive Value. Table 3Summary of studies evaluating CTP in seizures.ReferenceTypeSynopsisTimingFindings(Wiest, von Bredow et al., 2006)Case series15 patientsCTP/EEG 2 hVariable perfusion patterns: 3 subtle SE had hyperperfusion (2 EEG concordant); 12 postictal: 6 hypoperfusion with EEG focal slowing, 6 normal (with normal EEG)(Gelfand, Wintermark et al., 2010)Case series27 seizure patientsCTP/presentation 72 h37% had abnormal CTP; only 4% had hyperperfusion.Imaging within 2 h after seizure termination was associated with a higher likelihood of abnormal CTP.(Payabvash, Oswood et al., 2015)Case series18 seizure patientsSymptom/CTP 3 h14/18 patients (78%) had unilateral hyperperfusion. None had hypoperfusion(Van Cauwenberge, Dekeyzer et al., 2018)Case series133 seizure patients: 50% Todd’s phenomenon 28% postictal w/out Todd’s22% SE/recurrent seizuresSymptom/CTP1.9 hPatients with SE/recurrent seizures had higher incidence of hyperperfusion, 38% vs 11–16% (without or w/ Todd’s).CTP hyperperfusion had 38% sensitivity and 86% specificity for detecting ictal activity with an NPV of 83%.(Austein, Huhndorf et al., 2018)Case series37 with seizure-related stroke mimicsSymptom/CTP within 6 h, ¾ within 3 h60% hyperperfusion, 41% hypoperfusionHyperperfusion in thalamus or mesial temporal structures was significantly correlated with epileptiform EEG activity.On follow up MRI, only patients with CTP hyperperfusion demonstrated seizure-related signal abnormality in the thalamus or hippocampus.(Manganotti, Furlanis et al., 2019)Case series11 epileptic aphasic patientsSymptom/CTP within 4.5 hEEG within 24 h8/11 (73%) had CTP hyperperfusion and sharp waves on EEG.3/11 (27%) had hypoperfusion and slowing on EEG.(Lucas, Gariel et al., 2021)Case control45 seizure vs 50 ischemic strokeSymptom/CTP within 4.5 hEEG during hospitalizationHyperperfusion was more frequent in the seizure group compared to stroke group (36% vs 2%).Hyperperfusion had high specificity of 98% but weak sensitivity of 35% for diagnosing seizures.Specificity and PPV improved if the perfusion pattern was not restricted to a vascular territory(Gonzalez-Martinez, Trillo et al., 2021)Case series94 seizure and 21 SE, out of 284 S mimicsSymptom/CTP, mean 2.9 h50% of patients with seizures or SE had abnormal CTP.CBV was increased in 17% of seizure patients and 19% of SE.(Prodi, Danieli et al., 2022)Case series31 seizure patients out of 401 S/stroke mimicsn/aAbnormal CTP was present in 42% of seizure patients.23% hyperperfusion, 19% hypoperfusion.(Ameen Ahmad, Primiani et al., 2024)Case series103 patientsCTP/EEG within 6 h35 patients had clinical seizures: 1/3 had hyperperfusion, 1/3 hypoperfusion, 1/3 normal perfusion.CTP hyperperfusion had low sensitivity of 34% but high specificity 96% for identifying patients with clinical seizures.Abnormal CTP had moderate sensitivity of 69% for identifying clinical seizures.(Rapai Kasini, Payus et al., 2025)Case control31 seizures vs 31 LVO strokesSymptom/CTP within 48 hOnly 19% seizure patients had hyperperfusion; most had normal (39%) or hypoperfusion (36%).LVO: large vessel occlusion, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, PPV: Positive Predictive Value, NPV: Negative Predictive Value. Summary of studies evaluating CTP in seizures. LVO: large vessel occlusion, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, PPV: Positive Predictive Value, NPV: Negative Predictive Value. Wiest et al. attempted to investigate the utility of CTP in 15 patients who presented to the ER because of change in mental status or presumed witnessed epileptic seizure. (Wiest, von Bredow et al., 2006) Routine EEG was completed within 2 h of CTP. Three patients with subtle SE had hyperperfusion, two of which had EEG/CTP concordance with sharp waves, while one patient had a normal EEG. Twelve patients were deemed to be in the postictal state; half of them had hypoperfusion with concordant focal slowing on EEG, while the remaining six patients had normal perfusion and normal EEG. (Wiest, von Bredow et al., 2006) None of the patients had hyperperfusion. (Wiest, von Bredow et al., 2006) Another study found that 9 out of 27 patients presenting with seizures had CTP hypoperfusion, while only 1 patient had hyperperfusion. (Gelfand, Wintermark et al., 2010). However, Payabvash et al found that 14/18 (78%) patients presenting with seizure- related presentation (focal neurological deficits including aphasia, convulsions, or altered mental status) had unilateral hyperperfusion when scanned within 3 h of symptoms. Several of the patients had persistent neurological deficits raising the concern for undiagnosed NCSE in this cohort. (Payabvash, Oswood et al., 2015) Similarly, Restrepo-Vera et al. evaluated CTP findings in a cohort of 47 patients (27 patients with single seizure, 20 patients with SE). (Restrepo-Vera, Coscojuela et al., 2022) About 22% of patients with single seizures (diagnosed clinically) had hyperperfusion, compared to 60% of clinical SE, and 89% of electrographic SE. (Restrepo-Vera, Coscojuela et al., 2022) The lower rate of hyperperfusion in clinical SE compared with electrographic SE may be attributable to the timing of CTP, as some patients with clinical SE may have been scanned during the postictal period rather than during the ictal state. Our group showed that out of 35 patients who had clinical seizures and underwent CTP with a median of 2.6 h from onset of symptoms, about 1/3 had hyperperfusion, 1/3 had hypoperfusion, while 1/3 had a normal perfusion pattern.(Ameen Ahmad, Primiani et al., 2024) CT hyperperfusion demonstrated a low sensitivity of 34% and a high specificity of 96% for identifying clinical seizures. (Ameen Ahmad, Primiani et al., 2024) Similarly, in another recent study comparing 31 patients presenting with acute seizures vs. patients with acute stroke with large vessel occlusion, only 19% had hyperperfusion pattern. (Rapai Kasini, Payus et al., 2025) There were no differences in the CTP pattern between patients with first time seizure vs. epilepsy, patients with or without postictal deficits, focal seizures vs. generalized seizures, and among the various etiologies of seizures. (Rapai Kasini, Payus et al., 2025). Van Cauwenberge et al. analyzed 133 seizure patients who underwent CTP, classifying them into three groups: postictal with Todd’s phenomenon (n = 67), ictal (status epilepticus or recurrent seizures, n = 29), and postictal without Todd’s phenomenon (n = 37). In the ictal group, 38% demonstrated hyperperfusion, 52% had normal perfusion, and a minority exhibited hypoperfusion. Among patients with Todd’s phenomenon, 16% showed hyperperfusion, 21% hypoperfusion, and 63% normal perfusion, like the postictal group without Todd’s phenomenon (11% hyperperfusion, 19% hypoperfusion, 68% normal). Ictal patients had a higher incidence of hyperperfusion than postictal ones (p = 0.015). The sensitivity and specificity of hyperperfusion for detecting ictal activity were 38% and 86% respectively, with an NPV of 83%. (Van Cauwenberge, Dekeyzer et al., 2018). Lucas et al. compared CTP findings in 45 patients with seizures vs. 50 with ischemic stroke. Hyperperfusion was significantly more frequent in the seizure group (36% vs 2%, p < 0.01), with only one stroke patient displaying hyperperfusion. The presence of hyperperfusion on CTP had a high specificity (98%) and positive predictive value (94%) for diagnosing seizures, but limited sensitivity (35%) and moderate negative predictive value (63%). When hyperperfusion not restricted to a vascular territory was considered, specificity and PPV increased to 100%, while sensitivity and NPV were unchanged. Inter-observer agreement for identifying the hyperperfusion pattern was high. (Lucas, Gariel et al., 2021). In patients presenting with first-time seizure, CTP may have a role as well. (Kim, Lee et al., 2017) Out of 69 patients with first time seizures, 6 out of 7 had cortical and thalamic hyperperfusion. (Kim, Lee et al., 2017). In summary, regional CTP hyperperfusion has a high positive predictive value for diagnosing seizures, however, with low sensitivity. CT perfusion in epileptic stroke mimics About 17–25% of patients presenting with a concern for ischemic stroke have non-ischemic etiology, referred to as stroke mimics. (Hemmen et al., 2008, Mehta et al., 2014, Khalili et al., 2023) Seizures and status epilepticus comprise a sizable proportion of stroke mimics. In this regard, several studies have evaluated the utility of CTP in identifying patients with stroke mimics (mainly seizures and SE) comparing them to ischemic stroke. (Austein, Huhndorf et al., 2018). In a study of 43 patients with epileptic stroke mimics, CTP was abnormal in about 50%, with 13 (30%) patients having focal hyperperfusion and 8 (19%) had focal hypoperfusion. (Strambo, Rey et al., 2018) The cohort consisted of patients with either SE (13 patients, 30%) or seizures. Hyperperfusion was found in 9/13 (69%) with SE at any time (p < 0.01 if compared to patients without SE), in 4/5 (80%) for SE still clinically persistent at the moment of CTP and in 5/8 (62%) for a SE already resolved before CTP. Additionally, hyperperfusion was present in 4 patients with seizures but not SE. (Strambo, Rey et al., 2018). Restrepo-Vera et al. evaluated 47 patients initially suspected of stroke who were later diagnosed with either a single seizure (27 patients) or status epilepticus (20 patients). CTP was completed a median of 3 h from last seen asymptomatic. 51% of the patients had abnormal CTP (24 patients); 18 patients had CTP hyperperfusion, while six patients had focal hypoperfusion. CTP hyperperfusion was found in 38.3% of patients, most commonly in those with SE (60%, 12 patients) versus single seizures (22%, 6 patients). CTP hyperperfusion had a sensitivity of 60% and specificity of 77.8% for the detection of SE as a final diagnosis. All patients with hypoperfusion had a single seizure. Among those who had EEG (41 patients), nearly 89% of patients with SE on EEG had CTP hyperperfusion, compared to just 16% with non-epileptiform EEG and 42% of patients with interictal epileptiform discharges. Of note, the median time from last seen asymptomatic to obtaining EEG was 16 h, and from CTP to EEG was 8 h. (Restrepo-Vera, Coscojuela et al., 2022) In another study, 42% of seizure patients within a cohort of 401 individuals presenting with either stroke or stroke mimic exhibited abnormal CTP findings; among these, 23% demonstrated hyperperfusion, while 19% showed hypoperfusion. (Prodi, Danieli et al., 2022). Manganotti et al evaluated CTP in patients presenting with isolated aphasia due to either ischemic aphasia (27 patients) or stroke mimics (11 epileptic aphasia, 7 with other non-ischemic causes). (Manganotti, Furlanis et al., 2019) Out of 11 patients with epileptic aphasia, 8 presented with hyperperfusion pattern on CTP and sharp waves on EEG, while 3 patients showed slight hypoperfusion and slowing on EEG. The degree of hypoperfusion differed between patients with ischemic stroke and those with seizure-related hypoperfusion. In seizure-related cases, left hemispheric MTT was higher than the contralateral side, with asymmetry ranging from 11% to 22%, along with reduced CBF and CBV. In contrast, patients with ischemic aphasia exhibited asymmetry greater than 45%. (Manganotti, Furlanis et al., 2019). CT perfusion in Ictal-Interictal continuum patterns Patients with acute brain insults and critically ill patients often have abnormal EEG. (Kapinos, Trinka et al., 2018) About 35% to 41% of these critically ill patients have periodic and rhythmic patterns (PRPs) when monitored on continuous EEG. (Rodriguez Ruiz, Vlachy et al., 2017) Some of these PRPs lie on a spectrum between clear seizure activity and clear nonseizure activity, known as the ictal-interictal continuum. (Rodriguez Ruiz et al., 2017, Kapinos et al., 2018, Husari and Johnson, 2020). There are few reports correlating the findings of CTP with EEG patterns. Some of the earlier reports where EEG was done within 1 week of CTP, patients with epileptiform EEG had higher ipsilateral rCBF and rCBV and lower rMTT, compared to patients with slow or normal EEG. (Payabvash, Oswood et al., 2015) EEG slowing in the temporal lobe was observed significantly more frequently in patients with hyperperfusion (20/22; 90.9%) than in the hypoperfusion group. (Austein, Huhndorf et al., 2018) Hyperperfusion in the thalamus or mesiotemporal structures was significantly correlated with epileptiform EEG activity. (Austein, Huhndorf et al., 2018). In the cohort by G. Giovannini et al, 14/16 patients had definite SE while two patients had probable SE (according to Salzburg criteria). 12 out of 16 (75%) had a continuous ictal pattern, while four patients had intermittent recurrent focal seizures. All patients with continuous ictal pattern showed cortical hyperperfusion, while this was the case only for 3 out of six patients with intermittent recurrent seizures (p = 0.03) (one patient had thalamic but not cortical hyperperfusion). No differences in thalamic perfusion was found between continuous and intermittent patterns, although numbers were small. (Giovannini, Malagoli et al., 2021). In the study conducted in our center, CTP hyperperfusion was statistically significant with respect to diagnosing patients with either electrographic seizures or lateralized RPP (LPDs/LRDA) with a low sensitivity of 40% and a high specificity of 92%. (Ameen Ahmad, Primiani et al., 2024) Abnormal CTP (either hyper- or hypo- perfusion) had a relatively high sensitivity and high NPV in patients with electrographic seizures or lateralized RPP. In our cohort, 40% of patients with LPDs and 60% of patients with LRDA had CTP hyperperfusion. None of the patients with GPDs or GRDA had CTP hyperperfusion. (Ameen Ahmad, Primiani et al., 2024). A summary of the studies evaluating CTP in rhythmic and periodic patterns is presented in Table 4.Table 4Summary of studies evaluating CTP in RPP and electrographic seizures.ReferenceTypeSynopsisTimingFindings(Payabvash, Oswood et al., 2015)Case series13/18 patients with EEG within 1 week; 5 epileptiform, 6 non-epileptiform, 2 normalCTP/EEG within 1 weekPatients with epileptiform EEG had higher ipsilateral temporal/parietal rCBF/rCBV and lower temporal rMTT.(Gonzalez-Cuevas, Coscojuela et al., 2019)Case-controlOut of 19 SE patients:5 LPDsCTP/EEG mean ≈ 40 min in SE group4/5 patients with LPDs had regional cortical hyperperfusion(Spalletti, Pescini et al., 2020)Case reportA patient with a history of subdural hematoma, presenting with aphasiaCTP/EEG few hours1–2 Hz LRDA with intermixed sharp waves, CTP with concordant increased rCBF/rCBVTreated with lacosamide, with clinical and EEG improvement (resolution of LRDA).(Sheikh, Park et al., 2021)Case reportA patient with left subdural hematomaCTP/EEG few hours1 Hz left hemispheric LPDs + R; CTP showed increased left rCBF/rCBV, normal MTT; improved with ASM escalation.(Giovannini, Malagoli et al., 2021)Case series21 NCSE patients; 18 met Salzburg criteriaSymptom onset/CTP median 3 h4/6 patients with LPDs had cortical hyperperfusion, 1/6 had concomitant thalamic hyperperfusion.2/6 had normal CTPs.(Restrepo-Vera, Coscojuela et al., 2022)Case series27 seizure and 20 SE patients;Symptom/ CTP, median 3 hSymptom/EEG, median 16 hEEG/CTP, median 8 h90% of patients with electrographic/electroclinical SE had CTP hyperperfusion, compared to 16% of patients with non-epileptiform EEG and 42% of patients with interictal epileptiform discharges.55% of electrographic/electroclinical SE had thalamic hyperperfusion.(Ameen Ahmad, Primiani et al., 2024)Case series103 patients;8% had electrographic seizures and 31% RPPCTP/EEG within 6 h40% LPDs and 60% LRDA had hyperperfusion.No hyperperfusion in GPDs/GRDA.Hyperperfusion had 40% sensitivity, 92% specificity for electrographic seizures/lateralized RPP.ASM: anti-seizure medication, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, GPDs: generalized periodic discharges, GRDA: generalized rhythmic delta activity, RPP: rhythmic or periodic patterns, LPDs: lateralized periodic discharges, LRDA: lateralized rhythmic delta activity, rCBF: regional cerebral blood flow, rCBV: regional cerebral blood volume, rMTT: regional mean transit time, Table 4Summary of studies evaluating CTP in RPP and electrographic seizures.ReferenceTypeSynopsisTimingFindings(Payabvash, Oswood et al., 2015)Case series13/18 patients with EEG within 1 week; 5 epileptiform, 6 non-epileptiform, 2 normalCTP/EEG within 1 weekPatients with epileptiform EEG had higher ipsilateral temporal/parietal rCBF/rCBV and lower temporal rMTT.(Gonzalez-Cuevas, Coscojuela et al., 2019)Case-controlOut of 19 SE patients:5 LPDsCTP/EEG mean ≈ 40 min in SE group4/5 patients with LPDs had regional cortical hyperperfusion(Spalletti, Pescini et al., 2020)Case reportA patient with a history of subdural hematoma, presenting with aphasiaCTP/EEG few hours1–2 Hz LRDA with intermixed sharp waves, CTP with concordant increased rCBF/rCBVTreated with lacosamide, with clinical and EEG improvement (resolution of LRDA).(Sheikh, Park et al., 2021)Case reportA patient with left subdural hematomaCTP/EEG few hours1 Hz left hemispheric LPDs + R; CTP showed increased left rCBF/rCBV, normal MTT; improved with ASM escalation.(Giovannini, Malagoli et al., 2021)Case series21 NCSE patients; 18 met Salzburg criteriaSymptom onset/CTP median 3 h4/6 patients with LPDs had cortical hyperperfusion, 1/6 had concomitant thalamic hyperperfusion.2/6 had normal CTPs.(Restrepo-Vera, Coscojuela et al., 2022)Case series27 seizure and 20 SE patients;Symptom/ CTP, median 3 hSymptom/EEG, median 16 hEEG/CTP, median 8 h90% of patients with electrographic/electroclinical SE had CTP hyperperfusion, compared to 16% of patients with non-epileptiform EEG and 42% of patients with interictal epileptiform discharges.55% of electrographic/electroclinical SE had thalamic hyperperfusion.(Ameen Ahmad, Primiani et al., 2024)Case series103 patients;8% had electrographic seizures and 31% RPPCTP/EEG within 6 h40% LPDs and 60% LRDA had hyperperfusion.No hyperperfusion in GPDs/GRDA.Hyperperfusion had 40% sensitivity, 92% specificity for electrographic seizures/lateralized RPP.ASM: anti-seizure medication, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, GPDs: generalized periodic discharges, GRDA: generalized rhythmic delta activity, RPP: rhythmic or periodic patterns, LPDs: lateralized periodic discharges, LRDA: lateralized rhythmic delta activity, rCBF: regional cerebral blood flow, rCBV: regional cerebral blood volume, rMTT: regional mean transit time, Summary of studies evaluating CTP in RPP and electrographic seizures. ASM: anti-seizure medication, SE: status epilepticus, CTP: CT Perfusion, EEG: electroencephalogram, GPDs: generalized periodic discharges, GRDA: generalized rhythmic delta activity, RPP: rhythmic or periodic patterns, LPDs: lateralized periodic discharges, LRDA: lateralized rhythmic delta activity, rCBF: regional cerebral blood flow, rCBV: regional cerebral blood volume, rMTT: regional mean transit time, Patterns of CT perfusion changes The pattern of cerebral perfusion abnormalities differs between patients with seizures and those with stroke-related perfusion abnormalities. (Austein et al., 2018, Rapai Kasini et al., 2025). For epileptic patients with hyperperfusion, the most common pattern is unilateral cortical hyperperfusion sparing the deep white matter and subcortical structures (cortical ribbon pattern). (Austein et al., 2018, Van Cauwenberge et al., 2018, Rapai Kasini et al., 2025) The hyperperfusion crosses vascular territory boundaries. (Austein et al., 2018, Van Cauwenberge et al., 2018, Rapai Kasini et al., 2025) These two features help differentiate seizure-related abnormalities from patients with luxury perfusion following reperfusion of cerebral infarction. (Nagar, McKinney et al., 2009). In patients with seizures and hypoperfusion, most have cortical and subcortical hypoperfusion spanning multiple vascular territories, as compared to ischemic stroke hypoperfusion which respects the anatomical boundaries of vascular territories. This non-territorial pattern is also true of hyperpefusion seen in MR perfusion studies. (Yoo, Yun et al., 2017). Additionally, the degree of hypoperfusion differs. CTP in seizure-related hypoperfusion demonstrates mild decreases in CBF and CBV and mild increase in MTT, while stroke patients exhibited significantly higher Tmax and MTT ratios. (Austein et al., 2018, Lucas et al., 2021) The difference reflects underlying pathophysiology: deep hypoperfusion in stroke resulting from arterial occlusion and infarct core formation, whereas the more moderate hypoperfusion seen postictally is attributed to neuronal inhibition or regional hypoactivity. (Lucas, Gariel et al., 2021). In patients with post-stroke NCSE, hyperperfusion is typically seen outside the infarcted tissue with most cases having extra-lesional or peri-lesional hyperperfusion. (Merli, Romoli et al., 2025). CTP-EEG spatial localization There is typically a concordant spatial localization between CTP hyperperfusion and EEG ictal patterns. This is particularly true at a hemispheric and lobar level more so than the sub-lobar level. (Giovannini et al., 2021, Merli et al., 2023) Romoli et al. assessed CTP use in NCSE by using both conventional analysis and a novel reverse Tmax (rTmax) analysis. (Romoli, Merli et al., 2022) The authors defined hyperperfusion as regions with Tmax ≤ 3 s. They further stratified hyperperfused areas into subregions defined by Tmax thresholds of 3, 2, and 1 s. Both the standard and rTmax analyses had similar accuracy, but rTmax had superior spatial localization. rTmax 1 maps showed better sub-lobar localization with the EEG compared to standard CTP maps where the agreement was at the hemispheric level. (Romoli, Merli et al., 2022). Patients with lateralized rhythmic and periodic patterns typically exhibited concordance between the hemispheric lateralization of cerebral perfusion abnormalities and EEG findings. In cases with hyperperfusion, the laterality of CTP hyperperfusion consistently matched the lateralization of electrographic seizures, LPDs, and LRDA in all patients. Patients with hypoperfusion demonstrated concordant lateralization in approximately 80% of cases. (Ameen Ahmad, Primiani et al., 2024). MR perfusion MRI-related changes observed in patients experiencing recurrent seizures and status epilepticus (SE) are commonly termed peri-ictal magnetic resonance imaging (MRI) abnormalities (PMA). These abnormalities encompass alterations detected on diffusion-weighted imaging (DWI), fluid-attenuated inversion recovery (FLAIR), T2-weighted imaging (T2WI), contrast-enhanced imaging, and MR perfusion sequences. PMA can occur in up to 40–45% of SE patients. (Bosque Varela et al., 2023, Bosque-Varela et al., 2026) During SE, increased neuronal activity leads to elevated regional cerebral perfusion in an attempt to meet the heightened metabolic demand. This hyperperfusion can be detected using MRP imaging. If cerebral perfusion is no longer sufficient to match metabolic requirements, regional ischemia may occur, which can subsequently manifest as either cytotoxic edema—characterized by reduced diffusivity on DWI—or vasogenic edema, which appears as hyperintensity on FLAIR and T2-weighted sequences. (Daftari Besheli, Ahmed et al., 2022) These pathophysiological changes show good correlation with clinical imaging findings: MR perfusion abnormalities were observed in 37% of SE patients, followed by DWI changes in 27% and FLAIR changes in 18%. (Bosque Varela, Machegger et al., 2023). Given the scope of this review, we will primarily focus on MPR changes associated with seizure activity. For a comprehensive discussion of other PMA, readers are referred to several review articles and a recent metanalysis. (Mendes and Sampaio, 2016, Giovannini et al., 2018, Bosque-Varela et al., 2025). MR perfusion in patients with SE Compared to CTP, fewer studies have investigated the application of MR perfusion in patients with seizures and status epilepticus (SE). (Szabo et al., 2005, Kanazawa et al., 2015, Jaraba Armas et al., 2022) Both imaging modalities – DSC and ASL- have been used in patients with seizures, with ASL being more commonly used given lack of contrast administration and equal sensitivity. (Pizzini, Farace et al., 2013) A summary of studies evaluating MR perfusion in patients with SE and seizures is presented in Table 5.Table 5Summary of studies evaluating MR Perfusion in patients with SE and seizures.ReferenceTypeSynopsisTimingFindings(Szabo, Poepel et al., 2005)Case series10 focal SE; DSC-MRPEEG/MRP, mean 1.8 hAll patients had regional hyperperfusion on TTP, MTT, and CBV maps.Perfusion changes normalized on follow up MRI.(Pizzini, Farace et al., 2013)Case series19 patientsGroup A: high seizure frequency and < 5 h,Group B: low seizure frequency and < 28 hASL-MRP and DSC-MRPSymptom/MRP, mean 14 h<5h: 89% hyperperfusion; 11% had normal perfusion>5h: 70% hypoperfusion, 30% normalNo difference in perfusion patterns between ASL-MRP and DSC-MRP.(Kanazawa, Morioka et al., 2015)Case report2 patients with post-stroke NCSE; ASL-MRPWithin few hours of EEGBoth had regional hyperperfusion on ASL-MRP.(Jabeen, Cherukuri et al., 2017)Case seriesProspective study 30 SE/seizure patients with DWI5 underwent ASL-MRP.MRP within 24 h of admissionAll 5 patients had regional hyperperfusion, co-localized with DWI changes.(Shimogawa, Morioka et al., 2017)Case series15 patients with NCSEDWI, ASL-MRP, and EEGMRP and EEG obtained on day of admission73% had both ASL hyperperfusion and DWI hyperintensity27% had only ASL hyperperfusion.100% MRI and EEG concordanceSparing of precentral gyrus in all patients.(Yoo, Yun et al., 2017)Case series147 seizures, 17 seizure mimics; ASL-MRPSymptom/MRP divided < or > 5hIn the seizure group: 36% had hyperperfusion, 7% hypoperfusion.Hyperperfusion was more common in patients < 5 h vs > 5 h (58% vs 26%, p = 0.002)In the seizure mimic group, 1 patient had hypoperfusion, while 0 had hyperperfusion.(Ohtomo, Otsubo et al., 2021)Case series28 suspected NCSE (21 confirmed per EEG criteria); ASL-MRP/EEGMRP/EEG, mean 8 hAll 21 patients with NCSE had cortical hyperperfusion.3 patients without EEG confirmed NCSE had hyperperfusion.16/21 (76%) of NCSE had thalamic hyperperfusion.(Jaraba Armas, Sala-Padro et al., 2022)Case series19 aphasic SE5 had MR perfusion.2 had CTP.10 had SPECTn/a2/5 (40%) had MR hyperperfusion2/2 (100%) had CTP hyperperfusion8/10 (80%) had SPECT hyperperfusion(Kostner, Rebsamen et al., 2023)Case control51 patients; 18 ictal, 33 postictal.DCS-MRP/EEGSymptom/MRP, median 3.75 hPresentation/MRP, median 2 hHyperpefusion was present in 17/18 (94%) of ictal group vs. 0 in the post ictal group.73% of post ictal group had normal perfusion, 27% had hypoperfusion.DCS-MRP: Dynamic susceptibility contrast MR Perfusion, ASL-MRP: Arterial Spin Labeling-MR Perfusion, NCSE: non-convulsive status epilepticus, SE: status epilepticus, CTP: CT Perfusion, SPECT: Single-photon emission computed tomography, DWI: Diffusion-weighted imaging, rCBV: regional cerebral blood volume, rMTT: regional mean transit time, TTP: time-to-peak. Table 5Summary of studies evaluating MR Perfusion in patients with SE and seizures.ReferenceTypeSynopsisTimingFindings(Szabo, Poepel et al., 2005)Case series10 focal SE; DSC-MRPEEG/MRP, mean 1.8 hAll patients had regional hyperperfusion on TTP, MTT, and CBV maps.Perfusion changes normalized on follow up MRI.(Pizzini, Farace et al., 2013)Case series19 patientsGroup A: high seizure frequency and < 5 h,Group B: low seizure frequency and < 28 hASL-MRP and DSC-MRPSymptom/MRP, mean 14 h<5h: 89% hyperperfusion; 11% had normal perfusion>5h: 70% hypoperfusion, 30% normalNo difference in perfusion patterns between ASL-MRP and DSC-MRP.(Kanazawa, Morioka et al., 2015)Case report2 patients with post-stroke NCSE; ASL-MRPWithin few hours of EEGBoth had regional hyperperfusion on ASL-MRP.(Jabeen, Cherukuri et al., 2017)Case seriesProspective study 30 SE/seizure patients with DWI5 underwent ASL-MRP.MRP within 24 h of admissionAll 5 patients had regional hyperperfusion, co-localized with DWI changes.(Shimogawa, Morioka et al., 2017)Case series15 patients with NCSEDWI, ASL-MRP, and EEGMRP and EEG obtained on day of admission73% had both ASL hyperperfusion and DWI hyperintensity27% had only ASL hyperperfusion.100% MRI and EEG concordanceSparing of precentral gyrus in all patients.(Yoo, Yun et al., 2017)Case series147 seizures, 17 seizure mimics; ASL-MRPSymptom/MRP divided < or > 5hIn the seizure group: 36% had hyperperfusion, 7% hypoperfusion.Hyperperfusion was more common in patients < 5 h vs > 5 h (58% vs 26%, p = 0.002)In the seizure mimic group, 1 patient had hypoperfusion, while 0 had hyperperfusion.(Ohtomo, Otsubo et al., 2021)Case series28 suspected NCSE (21 confirmed per EEG criteria); ASL-MRP/EEGMRP/EEG, mean 8 hAll 21 patients with NCSE had cortical hyperperfusion.3 patients without EEG confirmed NCSE had hyperperfusion.16/21 (76%) of NCSE had thalamic hyperperfusion.(Jaraba Armas, Sala-Padro et al., 2022)Case series19 aphasic SE5 had MR perfusion.2 had CTP.10 had SPECTn/a2/5 (40%) had MR hyperperfusion2/2 (100%) had CTP hyperperfusion8/10 (80%) had SPECT hyperperfusion(Kostner, Rebsamen et al., 2023)Case control51 patients; 18 ictal, 33 postictal.DCS-MRP/EEGSymptom/MRP, median 3.75 hPresentation/MRP, median 2 hHyperpefusion was present in 17/18 (94%) of ictal group vs. 0 in the post ictal group.73% of post ictal group had normal perfusion, 27% had hypoperfusion.DCS-MRP: Dynamic susceptibility contrast MR Perfusion, ASL-MRP: Arterial Spin Labeling-MR Perfusion, NCSE: non-convulsive status epilepticus, SE: status epilepticus, CTP: CT Perfusion, SPECT: Single-photon emission computed tomography, DWI: Diffusion-weighted imaging, rCBV: regional cerebral blood volume, rMTT: regional mean transit time, TTP: time-to-peak. Summary of studies evaluating MR Perfusion in patients with SE and seizures. DCS-MRP: Dynamic susceptibility contrast MR Perfusion, ASL-MRP: Arterial Spin Labeling-MR Perfusion, NCSE: non-convulsive status epilepticus, SE: status epilepticus, CTP: CT Perfusion, SPECT: Single-photon emission computed tomography, DWI: Diffusion-weighted imaging, rCBV: regional cerebral blood volume, rMTT: regional mean transit time, TTP: time-to-peak. Szabo et al. reported 10 patients presenting to the ER with focal status epilepticus undergoing brain MR perfusion using dynamic susceptibility contrast (DSC) in addition to diffusion-weighted imaging (DWI). All patients demonstrated subtle but noticeable signs of hyperperfusion on TTP, MTT, and CBV maps. Similarly, Shimogawa reported 15 patients with NCSE all having ASL hyperperfusion, with most patients undergoing MRIs within few hours of onset of symptoms.(Shimogawa, Morioka et al., 2017). Ohtomo et al studied 28 patients with clinically suspected NCSE undergoing ASL MRI and EEG. (Ohtomo, Otsubo et al., 2021) Twenty-one patients had EEG confirmed NCSE (19 definite, 2 possible according to Salzburg criteria) while 7 did not have NCSE. Cerebral cortical hyperperfusion was present in 24 (86%) of 28 patients, while thalamic hyperperfusion was identified in 16 (57%) of 28 patients. Cerebral cortical hyperperfusion was detected in all 21 episodes with NCSE including two episodes with possible NCSE, leading to a sensitivity of 100.0% and specificity of 57.1%. Additionally, 3 of those without NCSE per EEG criteria had hyperperfusion as well. (Ohtomo, Otsubo et al., 2021) There was no difference between cortical hyperpefusion and thalamic hyperperfusion for diagnosing patients with NCSE. Kostner et al. conducted a case-control study comparing patients who underwent MRI-DSC during status epilepticus (SE) or active seizures with those scanned in the post-ictal period. (Kostner, Rebsamen et al., 2023) Regional hyperperfusion was observed in 94% (17/18) of patients in the ictal group, whereas no hyperperfusion was detected in the post-ictal group. Conversely, hypoperfusion was identified in only one patient (6%) in the ictal group compared to 27% in the post-ictal group (the rest had normal perfusion). CBF demonstrated superior sensitivity as a perfusion marker compared to CBV, with the most pronounced difference between groups observed in thalamic CBF measurements, followed by temporal lobe. (Kostner, Rebsamen et al., 2023). In a recent single-center, prospective cohort study of 223 patients with either SE or IIC who underwent MRI within 48 h of diagnosis, no significant difference was observed in the rate of PMA between groups. (Bosque-Varela, Machegger et al., 2026) PMA were identified in 23 of 49 patients (47%) in the IIC group compared with 64 of 149 patients (43%) in the SE group. Similarly, hyperperfusion rates did not differ significantly between groups, occurring in 41% of IIC patients and 38% of SE patients. Notably, low-frequency (0.5–1 Hz) periodic discharges or spike-and-wave patterns were more frequently associated with hyperperfusion compared with high-frequency periodic discharges/spike-and-wave patterns (>1–2.5 Hz) or LRDA. (Bosque-Varela, Machegger et al., 2026). However, in a prospective cohort of 20 patients with SE, 13 (65%) had focal hyperperfusion on ASL. (Matsuura, Maeda et al., 2015) Those patients with additional diffusion restriction (10 patients) had longer mean from onset to MRI acquisition (but with a large standard deviation) compared to those who had normal DWI. Subsequent ASL imaging revealed heterogeneity with most normalizing, whereas some demonstrated hypoperfusion or rarely persistent hyperperfusion. (Matsuura, Maeda et al., 2015). In a preliminary study of LGI-1 encephalitis patients, all six patients had ASL hyperperfusion and T2/FLAIR hyperintensity in the hippocampus and/or amygdala in a non-territorial distribution. These findings could either represent ictal phenomena or be due to the acute inflammation (Yedavalli, Hamam et al., 2022). MR perfusion in patients with seizures Like CTP, the timing of MRP relative to the seizure influences the perfusion pattern. In one study, patients who underwent MRP within 5 h of seizure onset were more likely to demonstrate hyperperfusion (8 out of 9 patients; 1 with normal perfusion), whereas those imaged within 28 h after seizure onset did not exhibit hyperperfusion (0 out of 10 patients); instead, 7 patients demonstrated hypoperfusion and 3 had normal perfusion. (Pizzini, Farace et al., 2013) This was concordant with EEG findings as well; those with hyperpefusion had more epileptiform EEG whereas the second group had slow activity. (Pizzini, Farace et al., 2013) A subsequent larger cohort study by Yoo et al had similar results regarding the 5-hour threshold, with hyperperfusion observed in 58% of patients imaged within 5 h versus 26% of those imaged later. (Yoo, Yun et al., 2017) Yoo et al examined ASL in 147 patients with seizures and 17 with seizure mimics. Among patients with seizures, 36% (53/147) demonstrated hyperperfusion and 7% (11/147) showed hypoperfusion. In the seizure mimic group, only one patient exhibited hypoperfusion, while none showed hyperperfusion. The incidence of hyperperfusion did not differ significantly by seizure type or semiology within the seizure cohort. (Yoo, Yun et al., 2017). MRP-EEG spatial localization In general, there is concordance between the area of MR hyperperfusion and EEG ictal pattern. (Szabo, Poepel et al., 2005) Additionally, there is typically concordance between areas of MR perfusion and diffusion restriction, although larger areas of hyperperfusion can be seen. (Szabo et al., 2005, Jabeen et al., 2017, Shimogawa et al., 2017) MRP changes can extend beyond the presumed ictal zone and most commonly involved cortical and subcortical areas involved in seizure generation and propagation including the hippocampus, ipsilateral thalamus (particularly the pulvinar nucleus), and contralateral cerebellum. (Matsuura, Maeda et al., 2015). Thalamic hyperperfusion Thalamic involvement is central in supporting and maintaining both focal and generalized seizures. (Szaflarski, 2021) Thalamic hyperperfusion on either CTP or MRP seems to have high specificity for status epilepticus while largely absent in patients with single seizures.(Giovannini et al., 2021, Ohtomo et al., 2021, Restrepo-Vera et al., 2022) However, only about 30–44% of published cohorts had thalamic hyperperfusion. (Giovannini et al., 2021, Restrepo-Vera et al., 2022, Romoli, Merli et al., 2022, Merli et al., 2025). Thalamic hyperperfusion seems to be specific for patients with SE compared to patients with single seizures or post ictal. (Restrepo-Vera et al., 2022, Kostner et al., 2023) In the cohort by Giovannini et al, 44% patients with definite NCSE had thalamic hyperperfusion compared to neither of the two patients with probable NCSE. However, there was no difference in thalamic involvement between patients with continuous ictal pattern or recurrent intermittent seizures. (Giovannini, Malagoli et al., 2021) Similarly, thalamic hyperperfusion on MRP was present in 16 of the 28 patients (57%), and in 76% of patients with either confirmed or possible NCSE, showing a sensitivity of 76.2% and a specificity of 100%. (Ohtomo, Otsubo et al., 2021). Thalamic hyperperfusion and Inter-Ictal continuum patterns Thalamic hyperperfusion can also be present in patients with IIC (either LPDs or LRDA). (Ohtomo, Otsubo et al., 2021) This finding suggests hyperexcitability of thalamocortical networks in patients with LPDs and LRDA, highlighting the involvement of several thalamic nuclei, particularly the pulvinar and dorsomedial nuclei, in the propagation of SE and modulation of the rhythmicity and periodicity of IIC patterns. (Katramados et al., 2009, Tsoures et al., 2017, Ohtomo et al., 2021). In the study by Ohtomo, thalamic hyperperfusion on MRP was present in most patients with RDA and periodic discharges, with a sensitivity of 92.3% − 93.3% and specificity of 73.3% − 84.6% respectively. Thalamic hyperperfusion was significantly associated with lateralized and bilateral independent RDA, without association with generalized RDA, EEG seizures, and brief rhythmic epileptiform discharges. Compared to cortical hyperperfusion, thalamic hyperperfusion had a significantly higher accuracy for diagnosing either LPDs or RDA. (Ohtomo, Otsubo et al., 2021) The underlying reason for this difference remains unclear but may be attributable to regional variations in blood flow and energy demand or to the circumscribed and relatively small structure of the thalamus compared to the cortex. Representative case illustrations Case 1 A 50-year-old presented with acute repetitive seizures with right hemi-body twitching and aphasia. He had several focal seizures in the ER. At time of CTP, the patient had dense right hemiparesis and aphasia with NIHSS of 13. He had additional focal seizures (left facial twitching and left gaze deviation) soon after the CTP was obtained. CTP (Fig. 1) shows left parietal and occipital hyperpefusion with ipsilateral thalamic hyperperfusion (long arrow for thalamic hyperperfusion and short arrows for cortical hyperperfusion).Fig. 1Case 1. CTP showing left parietal and occipital hyperperfusion with ipsilateral thalamic hyperperfusion (increase in CBV and CBF, and decrease in MTT and TMAX). Long arrow for thalamic hyperperfusion and short arrows for cortical hyperperfusion. Case 1. CTP showing left parietal and occipital hyperperfusion with ipsilateral thalamic hyperperfusion (increase in CBV and CBF, and decrease in MTT and TMAX). Long arrow for thalamic hyperperfusion and short arrows for cortical hyperperfusion. Case 2 Same patient in case 1 presented 2 months later following a seizure with right arm and leg weakness with aphasia. He received IV midazolam en route by EMS for right arm twitching. CTP (Fig. 2) obtained during the post ictal period showing regional hypo-perfusion with Tmax prolongation (arrows).Fig. 2Case 2. CTP with regional hypoperfusion in the left paroetal and occipital region seen in post-ictal state. Notice the decrease CBF and Tmax prolongation. Case 2. CTP with regional hypoperfusion in the left paroetal and occipital region seen in post-ictal state. Notice the decrease CBF and Tmax prolongation. Case 3 A 75-year-old presented with right gaze deviation, right facial droop, and right facial twitching. He had a tonic-clonic seizure in the ER and was admitted to neurology service. On day 2, he was alert but had persistent aphasia. CTP and EEG were ordered. CTP showed a small area of regional hyperperfusion in the left temporal region (arrows) (Fig. 3A), EEG (Fig. 3B) with left hemispheric (temporal max) LPDs with polyspike morphology at around 1–1.5 Hz (asterisk *).Fig. 3ACase 3. CTP showed a small area of regional hyperperfusion on CBF and CBV maps in the left temporal region (arrows) and decrease in Tmax and MTT (arrows).Fig. 3BCase 3. AP Bipolar EEG montage with left hemispheric (temporal max) LPDs with polyspike morphology at around 1-1.5 Hz (asterisk *). Case 3. CTP showed a small area of regional hyperperfusion on CBF and CBV maps in the left temporal region (arrows) and decrease in Tmax and MTT (arrows). Case 3. AP Bipolar EEG montage with left hemispheric (temporal max) LPDs with polyspike morphology at around 1-1.5 Hz (asterisk *). Case 4 A 70-year-old with history of TBI with right hemispheric encephalomalacia and post traumatic epilepsy, presented with focal SE of left facial rhythmic twitching with decrease consciousness. The patient was treated with several anti-seizure medications. EEG demonstrated persistent LPDs and left facial twitching (asterisk *) (Fig. 4A). CTP was obtained to guide treatment escalation. CTP showed persistent hyperperfusion (most noticed on CBV, arrows) (Fig. 4B). Ketamine was started after CTP findings.Fig. 4ACase 4. AP Bipolar EEG montage with right temporal-central LPDs and left facial twitching and corresponding EMG artifact in the left lateral chain (asterisk *).Fig. 4BCase 4. CTP images showing persistent hyperperfusion (most noticed as an increase in CBV, arrows). Case 4. AP Bipolar EEG montage with right temporal-central LPDs and left facial twitching and corresponding EMG artifact in the left lateral chain (asterisk *). Case 4. CTP images showing persistent hyperperfusion (most noticed as an increase in CBV, arrows). Case 5 A 30-year-old with metastatic mesothelioma with complex medical history who was admitted with encephalopathy and rhythmic twitching. EEG demonstrated focal refractory SE arising from right hemisphere (asterisk *) (Fig. 5A). Fig. 5B shows quantitative EEG demonstrating refractory focal SE with flame like pattern on FFT spectrogram (star) and rhythmicity spectrogram (arrow). MRP-ASL demonstrated hemispheric hyperperfusion (arrows). Additionally, MRP-ASL shows contralateral cerebellar hyper-perfusion (thick arrow) which is hypothesized to be due to the activation of corticopontocerebellar pathway during SE (Fig. 5C).Fig. 5ACase 5. AP Bipolar EEG monatgeb with ictal pattern over the right hemisphere with evolving periodic spike and overriding fast activity (asterisk *).Fig. 5BCase 5. Quantitative EEG demonstrating refractory focal SE with flame like pattern on FFT spectrogram (star) and rhythmicity spectrogram (arrow).Fig. 5CCase 5. MRP-ASL demonstrated hemispheric hyperperfusion (arrows) with contralateral cerebellar hyper-perfusion (thick arrow). Case 5. AP Bipolar EEG monatgeb with ictal pattern over the right hemisphere with evolving periodic spike and overriding fast activity (asterisk *). Case 5. Quantitative EEG demonstrating refractory focal SE with flame like pattern on FFT spectrogram (star) and rhythmicity spectrogram (arrow). Case 5. MRP-ASL demonstrated hemispheric hyperperfusion (arrows) with contralateral cerebellar hyper-perfusion (thick arrow). Timing of perfusion imaging As mentioned earlier, the timing of perfusion imaging critically affects sensitivity and specificity for detecting regional hyperperfusion in seizures and status epilepticus. The optimal imaging window appears to be within 3–6 h of symptom onset, with studies with CTP/MRP obtained later showing lower sensitivity (Giovannini et al., 2021, Merli et al., 2025)(Strambo, Rey et al., 2018). Role of CTP and MRP in clinical practice NCSE can occur following a convulsive seizure or convulsive SE, or it can arise without preceding clinical seizures. NCSE may be mistaken for a prolonged postictal state, encephalopathy, or stroke. Most patients present with impaired responsiveness with minimal clinical signs (Leitinger, Bosque-Varela et al., 2025). The diagnosis of NCSE remains challenging (Leitinger et al., 2015, Leitinger et al., 2023). Although EEG is critical in the diagnosis of SE, access to immediate EEG and interpretation of findings within the clinical context can be difficult. Furthermore, there is incomplete agreement, even among experienced electroencephalographers, regarding which EEG patterns are diagnostic of NCSE. Several criteria have been proposed for EEG diagnosis of NCSE. The most widely used and validated is the modified Salzburg Criteria (Leitinger et al., 2015, Leitinger et al., 2023). Perfusion imaging can aid in the diagnosis of NCSE. CTP or MRP hyperperfusion is highly specific for epileptic conditions, particularly when the hyperperfusion is cortically based involving multiple vascular territories. We recommend a comprehensive approach to the diagnosis of SE that incorporates clinical presentation, neurological examination, EEG, and perfusion imaging. Integrating all these elements will improve the accuracy of NCSE diagnosis and help avoid potential misdiagnosis and over- or under-treatment. Both CTP and ASL showed no significant differences in detecting hyperperfusion based on a recent meta-analysis (Bosque-Varela, Kuchukhidze et al., 2025). Decision to pursue either imaging modality depends on the clinical scenario. CTP offers a fast method to aid in the diagnosis of status epilepticus and other epileptic conditions particularly in the emergency department. MRP offers the added benefit of obtaining structural sequences which will provide insight into the etiology of seizures/status epilepticus. Additionally, MRP-ASL doesn’t use contrast agent which might be beneficial in certain patients along with no risk of radiation. Conclusions and future directions . CTP or MRP hyperperfusion is highly specific for epileptic conditions, particularly when the hyperperfusion is cortically based involving multiple vascular territories. Thalamic hyperperfusion is specific for SE. Furthermore, CTP or MRP has been found to have relatively high sensitivity and specificity for diagnosing status epilepticus. However, the accuracy in patients with seizures and IIC patterns can be variable, largely due to the timing of the scan relative to the onset of seizure activity. Hence, a normal perfusion scan does not exclude the diagnosis of status epilepticus or recent seizure activity. Currently, CTP is useful in diagnosing patients presenting with status epilepticus or acute repetitive seizures with high specificity and moderate sensitivity. It can help identify patients either in ictal or peri-ictal state vs post-ictal state. However, future studies are needed to prospectively evaluate the utility of perfusion imaging in patients with RPP and IIC EEG patterns. The findings of such have the potential to guide diagnosis and management of such patients. Disclosures: None of the other authors have relevant conflicts of interest to be disclosed. Financial/Funding support None.