Time to detect periictal hyperperfusion following short acute symptomatic seizures using 1.5-Tesla arterial spin labeling magnetic resonance perfusion imaging

Abstract

Background: We evaluated the usefulness of capturing periictal hyperperfusion for the pathophysiological diagnosis of acute symptomatic seizures (ASS) using 1.5-Tesla (T) arterial spin labeling (ASL) perfusion images and examined the relationship between the time from ASS cessation to ASL imaging and the visualization of periictal hyperperfusion. Patients & Methods: In four patients who presented short ASS, we retrospectively analyzed the performance status and findings of 1.5-T ASL with triple post-labeling delays (PLDs) of 1.5, 1.75 and 2.0 s, as well as routine electroencephalography (EEG). Results: In two patients where ASL imaging was performed 1 or 9 h after ASS, periictal ASL hyperperfusion was markedly visualized. In one patient where images were taken 11 h later, fairly good visualization was obtained. The increase in signal intensity peaked at a PLD of 1.5 s and gradually attenuated with PLDs of 1.75 and 2.0 s. However, the areas where the signal remained intense even at a PLD of 2.0 s had a strong anatomical relationship with the lesion. No clear periictal hyperperfusion was visualized on ASL images taken 13 h later. Although paroxysmal discharges were recorded in one patient where EEG was performed 40 min after ASS, no paroxysms were detected in the other three patients whose EEG was recorded 8 h to 2 days later. Conclusion: We consider it appropriate to first perform ASL within 11 h, and then verify the results with subsequent EEG to accurately diagnose the pathophysiology of ASS.

Introduction

Seizures are classified into early seizures, also referred to as acute symptomatic seizures (ASS) which occur within 7 days of the onset of acute lesions such as stroke, and late seizures, which occur later. While the former is caused by the breakdown of the blood-brain barrier due to lesions, local metabolic changes, and direct stimulation of the cerebral cortex by blood breakdown products, the latter is thought to occur when the organization mechanism begins and epileptogenic foci are formed due to cortical gliosis. The pathophysiological mechanisms of these two types of seizures are thought to be different[1]. However, in recent years, the CAVE score[2] for intracerebral hemorrhage and the SeLECT score[3] for cerebral infarction have been commonly used to predict the onset of late seizures after stroke, with the presence of ASS as a key factor in both scores. In addition, there are two types of ASS: short ASS, characterized by a single convulsive seizure, and acute symptomatic status epilepticus, characterized by either convulsive or non-convulsive status epilepticus. The latter is thought to occur more frequently as a late seizure[4]. Therefore, in a neurological emergency setting, a prompt and accurate pathophysiological diagnosis of ASS is crucial to effectively guide to subsequent treatments.

The pathophysiological diagnosis of ASS, which is similar to that of structural focal epilepsy, traditionally relies on routine electroencephalography (EEG). However, this approach has limitations due to its availability only during weekday hours in most acute hospitals in Japan, including ours[5,6,7,8,9,10]. To compensate for this drawback, it has been reported that combining arterial spin labeling (ASL) perfusion images with magnetic resonance imaging (MRI) examinations, which can be performed 24 h a day with recent advancements in the treatment of acute cerebral infarction, can aid in the pathophysiological diagnosis of ASS by capturing periictal (or ictal) ASL hyperperfusion[11,12,13]. However, most of these studies have been conducted using pseudocontinuous labeling ASL with a 3-Tesla (T) MRI machine, while studies using the more widely used 1.5-T ASL remain limited as it is less effective in labeling, resulting in poorer spatial resolution of ASL images and increased contamination from intravascular signals, i.e., arterial transit artifacts (ATA)[10, 14, 15].

We have emphasized the importance of creating fusion ASL images superimposed onto anatomical MRI, such as diffusion-weighted images (DWI) and fluid-attenuated inversion recovery images (FLAIR), to improve the spatial resolution of ASL images when evaluating them, since diagnosis with 1.5-T ASL is often difficult using only the original pseudocolorized ASL images. By doing this, ASL signals in the subarachnoid space can be identified as ATAs[7,8,9,10, 16,17,18,19,20,21]. In addition, we have reported that subtraction of interictal (or periictal)-interictal arterial spin labeling co-registered to conventional MRI (SIACOM) is also useful in eliminating background signals and ATAs to clarify periictal hyperperfusion clearer[7,8,9,10, 14, 16,17,18,19,20,21].

We further reported the importance of evaluating the hemodynamic state of periictal hyperperfusion by adding images with a slower post-labeling delay (PLD) to the conventional PLD of 1.5 s, leveraging the sensitivity of ASL to blood flow velocity, that is, arterial transit time[22]. On ASL with 3-T pseudocontinuous labeling, we used a slow PLD of 2.5 s based on angiographic findings[22,23,24,25,26,27]; however, on 1.5-T ASL, we chose 2.0 s as a slow PLD, as imaging with a PLD that is too slow is prone to rapid T1 decay[14, 15]. To evaluate the hemodynamic state in more detail, we performed ASL imaging using triple PLDs, which included an additional PLD of 1.75 s[7,8,9,10, 16,17,18,19,20,21]. Periictal hyperperfusion generally demonstrates fast hemodynamics[22], and therefore exhibits maximum signal intensity at a PLD of 1.5 s, with the signal intensity decreasing as the PLD slows, in addition to the effect of rapid T1 decay. However, we have previously reported that the area in which a signal persists at a PLD of 2.0 s can be identified as the area with the greatest increase in the blood flow, i.e., the most epileptically activated area[7,8,9,10, 15, 18,19,20].

Furthermore, the detection of periictal hyperperfusion has been reported to be dependent on the intensity and duration of the original seizure, in addition to the time from seizure cessation to imaging[5, 8, 9, 12, 14, 16, 17, 20,21,22, 26, 28]. On the other hand, the periictal hyperperfusion recording time remains unclear, as there is currently no existing method to quantitatively evaluate the former two factors. We retrospectively analyzed the performance status and findings of 1.5-T MRI with ASL and EEG performed during the periictal state in four patients with short ASS, and reviewed the usefulness and limitations of both in diagnosing the pathophysiology of ASS. ASL imaging was conducted with triple PLDs and evaluated using three images: original ASL images, fusion ASL images, and SIACOM. In particular, we examined the relationship between the time from ASS cessation to ASL imaging and the visualization of periictal hyperperfusion.

Patients and methods

Patients

The study included four patients with short ASS (three males and one female) treated over a 2-year period from July 2022 to June 2024. During this time, EEG and 1.5-T MRI with ASL were dually performed by the neurological emergency department of our hospital. The patients’ ages ranged from 53 to 88 years (mean: 76.3 years). Ethical approval was obtained from the Institutional Review Board of Hachisuga Hospital (No. 22-1), which waived the requirement for written informed consent due to the retrospective study design.

Methods

Arterial Spin Labeling

ASL was performed as part of a routine MRI examination. Pseudocontinuous ASL was conducted using a 1.5-T scanner (ECHELON OVAL V6; FUJIFILM Healthcare, Tokyo, Japan) equipped with a 15-channel receive-only head coil for signal reception, as previously described[7,8,9,10, 16,17,18,19,20,21]. ASL was performed using a three-dimensional gradient and spin-echo sequence with background suppression for perfusion imaging of the entire brain. The acquisition parameters were as follows: phase encoding in the z-direction, 28; time to repeat, 4291 ms; echo time, 17.4 ms; field of view, 250 mm; matrix, 128 × 128; slice thickness, 6 mm; reconstruction interval, 3 mm; and number of slices, 48. The labeling duration was 1.5 s. Three PLDs of 1.5 s, 1.75 s, and 2.0 s were selected. The ASL acquisition time for each PLD was 3 min and 1 s respectively.

The obtained ASL images were pseudocolorized and used as the original ASL images. To create the fusion images, the original ASL images were superimposed onto the anatomical MRI, including DWI and FLAIR, using a SYNAPSE VINCENT volume analyzer (FUJIFILM Healthcare, Tokyo, Japan), as previously described[7,8,9,10, 16,17,18,19,20,21].

Interictal ASL was performed in all patients at 13 days–3 months after the admission. The interictal state was confirmed using EEG performed on the same day. The SIACOM was created using these images; the data from each voxel of the ictal ASL was automatically subtracted from the data of the corresponding voxel of the interictal ASL, and then superimposed onto the same anatomical MRI that created the fusion images using SYNAPSE VINCENT, as previously described[7,8,9,10, 16,17,18,19,20,21].

ASL findings were evaluated by visual inspection by board-certified neurosurgeons (S.I. and T.S.) experienced in ASL interpretation and blinded to the clinical data, as described in our previous reports[7,8,9,10, 16,17,18,19,20,21]. There were no differences observed in their independent assessments.

Electroencephalography

Routine EEG recordings were obtained during the day on weekdays using a digital EEG machine (Neurofax 1214; Nihon-Kohden, Tokyo, Japan) with electrode placement performed according to the International EEG 10-20 system. EEG findings were evaluated by visual inspection by two board-certified electroencephalographers (A.S. and H.S.) who were blinded to the clinical data. There were no differences recorded in the electroencephalographer’s interpretations and independent assessments. The EEG patterns were classified according to the American Clinical Neurophysiology Society’s Standardized Critical Care EEG Terminology, published in 2021 (ACNS 2021)[29].

Results (Table 1)

Table 1. Clinical, magnetic resonance imaging (MRI) including arterial spin labeling (ASL) perfusion image, and elctroencephalographic (EEG) findings of 4 patients with acute symptomatic seizure

Patient No.			1	2	3	4
Age and Gender			88M	76M (Lt handed)	53M	88F
Acute lesion (Volume of ICH)			Acute bleeding in Lt chronic subdural hematoma	Subcortical hemorrhage in Rt temporo-parietal lobe (10.8 mL)	Subcortical hemorrhage in Lt temporo-parietal lobe (2.8 mL)	Acute subdural hematoma, Rt
Symptoms at onset			Impaired awareness (FIAS?)	Lt hemiparesis and aphasia	Sensory aphasia	Headache and nausea
Timing of admission			Out of consult. hours	Within consult. hours	Out of consult. Hours	Out of consult. Hours
Seizure			FBTCS with Rt conjugate deviation at admission	FBTCS from Lt upper limb at admission	FBTCS with Rt conjugate deviation at admission, followed by restless confusion	FBTCS with Lt conjugate deviation 12 h after admission
ASM at seizure onset			DZP 5 mg iv→fPHT 750 mg div	DZP 5 mg iv→LCM 100mg div	DZP 5 mg iv→MDZ div	DZP 5 mg iv→PER 2mg div
Periictal examination
	Initial examination		MRI	EEG	MRI	EEG
	MRI
		Interval with seizure	1 h	9 h	11 h	13 h
		Detected increased ASL signals on
		Original ASL images	Markedly	Markedly	Fairly	-
		Fusion ASL images	Markedly	Markedly	Fairly	-
		SIACOM	Markedly	Markedly	Fairly	-
		Decreased signals on neighboring cortex	+	-	-	-
		Hyperintensity on DWI	-	-	-	-
	EEG
		Interval with seizure	2 days	40 min	13 h	8 h
		Findings	No paroxysm Intermittent slow wave at Lt fronto-temporal region	0.25-0.35 Hz LPDs at Rt parietal region	No paroxysm Intermittent slow waves at Lt temporal region	No paroxysm Slowing at Rt hemisphere
Interictal examination
	Interval with arrival		33 days	3 m	13 days	16 days
	EEG findings		No paroxysm	No paroxysm Intermittent slow waves, Rt. Hemisphere, especially central region	PPDA at Lt temporal region	PPDA at Rt posterior quadrant
ASM at outpatient room			LEV 1000mg po (with patient’s wish)	LCM 200mg po (CAVE score 3 points (1,0,1,1))	LEV 1000→750mg po (CAVE score 3 points (1,1,0,1))	PER 2mg→LEV 500mg po
Follow-up			No seizure at 1 y 9 m	No seizure at 10 m	No seizure at 1 y 1 m	No seizure at 3 m

Abbreviations: M, male; F, female, Rt, right; Lt, left; ICH, intracerebral hemorrhage; FIAS, focal impaired awareness seizure; h, hour; FBTCS, focal to bilateral tonic-clonic seizure; ASM, anti-seizure medication; DZP, diazepam; LCM, lacosamide; MDZ, midazolam; fPHT, fosphenytoin; PLD, post-labeling delay; LPD, lateralized periodic discharge; SIACOM, subtraction of ictal(periictal)-interictal ASL image co-registered to conventional MRI[14]; PPDA; persistent polymorphous delta activity; LEV, levetiracetam; CAVE score[2]; 1 point for each of cortical involvement, age<65 years, volume>10 mL, and early seizures within 7 days for intracerebral hemorrhage.

Arterial spin labeling (ASL)

Patient 1 had thin, chronic subdural hematomas on both sides (Fig. 1A), and a lesion highly suggestive of fresh bleeding in the hematoma cavity of the left frontal region (Fig. 1B, red arrows). MRI with ASL was performed 1 h after a generalized seizure that occurred during out-of-hours transport, with the seizure being terminated with an intravenous injection of 5mg of diazepam (DZP). This patient had been previously reported as Patient 2 in our previous report[8]; therefore, the ASL images showed different slice planes. Original ASL images at a PLD of 1.5 s revealed a marked increase in signal intensity in the cortex surrounding the fresh bleeding from the left frontal cortex to the tip of the temporal lobe (Fig. 1C, white arrows). This increase in signal intensity gradually diminished with PLDs of 1.75 and 2.0 s (Fig. 1D and E); however, signal intensity at the cortex surrounding the fresh bleeding persisted even at a PLD of 2.0 s (Fig. 1E, white arrows). In addition, the left posterior quadrant, posterior to the area of ​​increased signal intensity, demonstrated decreased signal intensity (Fig. 1C, dotted arrows). Fusion ASL images with DWI showed a strong anatomical relationship between the fresh bleeding (Fig. 1F–H, red arrows), the area of markedly ​​increased signal intensity at a PLD of 1.5 s (Fig. 1F white arrows) and the area of ​​residual signal intensity at a PLD of 2.0 s (Fig. 1H, white arrows). Furthermore, most areas of increased signal intensity in the contralateral right hemisphere corresponded to the sulci, which were determined to be the ATA of the distal segment of the middle cerebral artery (MCA). On SIACOM, most of the ATA and background signals were eliminated, and the periictal hyperperfusion was more clearly depicted (Fig. 1I–K). No cortical hyperintensity was observed on DWI at the area of ​​increased blood flow on the ASL images (Fig. 1B).

Fig. 1.

Patient 1. CT scan (A), DWI (B), (C–E) original ASL images at PLDs of 1.5 s (C), 1.75 s (D) and 2.0 s (E), (F–H) fusion ASL images with DWI at PLDs of 1.5 s (F), 1.75 s (G) and 2.0 s (H), and (I-K) SIACOM with DWI at PLDs of 1.5 s (I), 1.75 s (J) and 2.0 s (K), performed 1 h after ASS. Refer to the text for details.

CT, computed tomography; DWI, diffusion-weighted imaging; ASL, arterial spin labeling; PLD, post-labeling delay; SIACOM, subtraction of ictal (or periictal)-interictal ASL images co-registered with conventional magnetic resonance images; ASS, acute symptomatic seizure.

Patient 2 developed left hemiparesis and aphasia and was transferred to our hospital. On arrival, the patient developed a generalized convulsion beginning in the left upper limb. After halting the seizure with an intravenous injection of 5 mg of DZP, a computed tomography (CT) scan revealed approximately 10.8 mL of subcortical hemorrhage in the right temporo-parietal lobe (Fig. 2A). MRI performed 9 h later confirmed that the hematoma had not increased over time (Fig. 2B). A marked increase in signal intensity in the cortex of the right fronto-temporo-parietal lobe surrounding the hematoma cavity (Fig. 2C, red arrows) was evident on the original ASL images at a PLD of 1.5 s (Fig. 2C, white arrows). This increase in signal intensity gradually attenuated at PLDs of 1.75 and 2.0 s (Fig. 2D and E), but persisted at the area surrounding the hematoma even at PLD of 2.0 s (Fig. 2E, white arrows). There was no decrease in the signal intensity in any of the areas close to increased signal intensity. Fusion ASL images with FLAIR showed a strong anatomical relationship between the hematoma and the areas of increased signal intensity at a PLD of 1.5 s and the areas of residual high signal intensity at a PLD of 2.0 s (Fig. 2F–H). In addition, most of the areas with increased signal intensity in the contralateral left hemisphere were identified as the ATA of the distal MCA, as they coincided with the sulci. On SIACOM, most of the ATA and background signals were eliminated, and the periictal hyperperfusion was more clearly depicted (Fig. 2I–K). No cortical hyperintensity was observed on DWI at the area of ​​increased blood flow on the ASL images (Fig. 2L)

Fig. 2.

Patient 2. CT scan (A), FLAIR (B), (C–E) original ASL images at PLDs of 1.5 s (C), 1.75 s (D) and 2.0 s (E), (F–H) fusion ASL images with FLAIR at PLDs of 1.5 s (F), 1.75 s (G) and 2.0 s (H), (I–K) SIACOM at PLDs of 1.5 s (I), 1.75 s (J) and 2.0 s (K) with FLAIR, and DWI (L), performed 9 h after ASS. (M-N) EEG (M) and voltage topography of periodic discharge(N), performed 40 min after ASS. Refer to the text for details.

CT, computed tomography; FLAIR, fluid-attenuated inversion recovery; ASL, arterial spin labeling; DWI, diffusion-weighted imaging; PLD, post-labeling delay; SIACOM, subtraction of ictal (or periictal)-interictal ASL images co-registered with conventional magnetic resonance images; ASS, acute symptomatic seizure; EEG, electroencephalography.

Patient 3 was transferred to our hospital due to sensory aphasia. Upon admission, the patient experienced a generalized seizure with right conjugate deviation. After the generalized seizure was stopped by intravenous administration of 5 mg of DZP, CT scan revealed approximately 2.8 mL of subcortical hemorrhage in the left temporo-parietal lobe (Fig. 3A). MRI performed 11 h after the onset of the seizure confirmed that the hematoma had not increased in size over time (Fig. 3B). Original ASL images at a PLD of 1.5 s showed a slight increase in the signal intensity at the cortex of the left temporo-occipital lobe (Fig. 3C, white arrows), immediately above and posterior to the hematoma cavity (Fig. 3C, white dotted arrows). This increase in signal intensity was gradually attenuated with PLDs of 1.75 and 2.0 s (Fig. 3D and E), and remained at the same area even at a PLD of 2.0 s (Fig. 3E, white arrows). Decreased signal intensity was not observed in any of the areas close to increased signal intensity. Fusion ASL images with FLAIR showed a strong anatomical relationship between the hematoma and the area of increased signal intensity at a PLD of 1.5 s and the area of residual signal at a PLD of 2.0 s (Fig. 3F–H). SIACOM eliminated most of the background signals and ATA, and the periictal hyperperfusion was more clearly observed (Fig. 3I-K). No cortical hyperintensity was observed on DWI at the area of ​​increased blood flow on the ASL images (Fig. 3L).

Fig. 3.

Patient 3. CT scan (A), FLAIR (B), (C–E) original ASL images at PLDs of 1.5 s (C), 1.75 s (D) and 2.0 s (E), (F–H) fusion ASL images with FLAIR at PLDs of 1.5 s (F), 1.75 s (G) and 2.0 s (H), and (I–K) SIACOM at PLDs of 1.5 s (I), 1.75 s (J) and 2.0 s (K) with FLAIR, DWI (L), performed 11 h after ASS. Refer to the text for details.

CT, computed tomography; FLAIR, fluid-attenuated inversion recovery; ASL, arterial spin labeling; PLD, post-labeling delay; SIACOM, subtraction of ictal (or periictal)-interictal ASL images co-registered with conventional magnetic resonance images; DWI, diffusion-weighted imaging; ASS, acute symptomatic seizure.

Patient 4 was found at home collapsed, and was transported to the hospital by an ambulance. CT scan revealed a thin, acute subdural hematoma in the right posterior quadrant (Fig. 4A, red arrows). He had a generalized convulsive seizure with left conjugate deviation 12 h after admission, which was stopped by intravenous administration of 5 mg of DZP. MRI performed 13 h after the seizure revealed no increase in hematoma size (Fig. 4B). The original ASL images at a PLD of 1.5 s revealed a slight increase in signal intensity in the cortex immediately below the hematoma cavity (Fig. 4C, white arrows); however, this was considered negative because the degree of signal attenuation at PLDs of 1.75 and 2.0 s was not different from that in the contralateral hemisphere (Fig. 4D and E), and no clear increase in signal intensity was observed in the same area on fusion ASL images with FLAIR (Fig. 4F–H) and SIACOM (Fig. 4I–K). No cortical hyperintensity was observed on DWI.

Fig. 4.

Patient 4. CT scan (A), FLAIR (B), (C–E) original ASL images at PLDs of 1.5 s (C), 1.75 s (D) and 2.0 s (E), (F–H) fusion ASL images with FLAIR at PLDs of 1.5 s (F), 1.75 s (G) and 2.0 s (H), and (I–K) SIACOM at PLDs of 1.5 s (I), 1.75 s (J) and 2.0 s (K) with FLAIR, performed 13 h after ASS. Refer to the text for details.

CT, computed tomography; FLAIR, fluid-attenuated inversion recovery; ASL, arterial spin labeling; PLD, post-labeling delay; SIACOM, subtraction of ictal (or periictal)-interictal ASL images co-registered with conventional magnetic resonance images; ASS, acute symptomatic seizure.

Electroencephalography (EEG)

In Patient 2, who was transported within the consultation time, EEG was performed 40 min after ASS, and localized periodic discharges (PDs) of 0.25–0.35 Hz were observed mainly in the right parietal region (Fig. 2 M, N). In the other three patients, the transport was outside the consultation time, and EEG could not be performed immediately. In Patient 4, EEG was performed 8 h after ASS, and irregular slow waves were observed almost continuously in the right hemisphere of the affected side; however, no paroxysmal discharges were recorded. In Patients 1 and 3, EEG was performed 2 days and 13 h later, respectively. Intermittent slow waves were observed in the left fronto-temporal and left temporal regions, respectively, where the lesions were located; however, no paroxysmal discharges were recorded.

Seizure outcome

Patient 1 continued receiving 100 mg of levetiracetam (LEV) after discussion with the patient. One year and 9 months have passed since the ASS, and no subsequent seizures have been reported. Patients 2 and 3 both had subcortical hemorrhage, and their CAVE scores (1 point each for cortical involvement, age < 65 years, volume > 10mL, and early seizure)[2] were calculated as 3 points each (1,0,1,1 for Patient 2 and 1,1,0,1 for Patient 3, respectively). Therefore, both continued to receive 200 mg of lacosamide and 750 mg of LEV, following discussions with the patients. 10 months and 1 year 1 month have passed since the ASS respectively for these patients, with no additional seizures reported. Patient 4 became drowsy after oral administration of 2 mg of perampanel. As a result, the antiseizure medication (ASM) was changed to 500 mg of LEV. Three months have passed since the onset of ASS, and no seizures have occurred.

Discussion

The lesions responsible for the short ASS in the four patients in this study were hemorrhagic lesions that did not require emergency surgery. The ASS lasted for several minutes and was focal to bilateral tonic-clonic seizures stopped by intravenous administration of 5 mg of DZP. In addition, because no cortical hyperintensity on DWI was observed at the area of ​​increased blood flow on ASL, it was not a prolonged seizure or status epilepticus that would cause cytotoxic edema due to the uncoupling of metabolism and blood flow[5, 6, 22, 30, 31]. Therefore, the intensity and duration of the seizures in the four patients were probably similar. We believe that the time from ASS to ASL imaging played a major role in the visualization of periictal hyperperfusion in this case.

In the present study, periictal hyperperfusion was markedly visualized in all ASL images of Patients 1 and 2, where ASL images were taken at 1 and 9 h after ASS, respectively. Periictal hyperperfusion was fairly well visualized in Patient 3, who underwent imaging 11 h later. However, in Patient 4, no apparent periictal hyperperfusion was visualized on the ASL images taken 13 h later. Although the number of cases was limited (four) and further investigation is required, it is likely that periictal hyperperfusion can be visualized even with a 1.5-T ASL images taken within approximately 11 h following a short ASS.

The longer the time until imaging, the greater the attenuation in the signal intensity of periictal hyperperfusion. We previously reported that significant periictal hyperperfusion is accompanied by decreased blood flow in the vicinity due to blood steal[14, 22, 28]; however, this was only observed in Patient 1 in this study, and not in Patients 2 and 3. It is likely that the degree of increase in blood flow decreased after several hours prior to imaging, whereas blood stealing from nearby areas improved accordingly.

The hemodynamics of periictal hyperperfusion in the present study, similar to those of structural focal epilepsy[7,8,9,10, 19, 20], peaked at a PLD of 1.5 s and gradually attenuated at PLDs of 1.75 and 2.0 s, with no change in imaging time. Fusion ASL images and SIACOM clearly depicted the strong anatomical relationship with areas where intense signals persisted even at a PLD of 2.0 s and the acute lesions, and were extremely useful in diagnosing the pathophysiology of ASS.

In our previous report, we performed MRI with ASL at an average of 2.8 h following seizures in eight patients with dementia-related epilepsy[8], and within 1 h in all five patients with post-intracerebral hemorrhage epilepsy[9]. However, in the four patients reported here, MRI with ASL was delayed by several hours with the exception of Patient 1. The likely cause is that the primary focus of subsequent imaging shifted to the monitoring the expansion of hemorrhagic lesions over time, in the case hemorrhagic lesions were seen on CT scans at the time of transport and emergency surgery was not immediately performed.

EEG was performed 40 min after ASS in Patient 2, who was transported within the consultation time, and localized PDs of 0.25–0.35 Hz were recorded mainly in the right parietal region. Based on ACNS 2021[29], which provides guidelines for interpreting the findings of continuous EEG monitoring, it was determined that the patient was in the interictal-interictal continuum of ASS due to a subcortical hemorrhage in the right parietal lobe. However, in the other three patients, who were transported outside of the consultation time, EEG was recorded 8 h to 2 days later with no paroxysmal discharges detected. With a delay in performing the EEG examinations, the sensitivity of detecting paroxysmal discharges decreases. Therefore, we believe that it is often challenging to diagnose the pathophysiology of ASS using EEG alone, which cannot be performed 24 hours a day, similar to the diagnosis of structural focal epilepsy[7,8,9,10]. Currently, it is standard procedure for us to first perform MRI with ASL, and then verify the results with a subsequent EEG to accurately diagnose the pathophysiology of ASS[7,8,9,10].

There is currently no existing guideline on how long ASM should be continued for patients with short ASS. Therfore, we often refer to the CAVE[2] and SeLECT scores[3] in patients with stroke, consider factors such as the presence or absence of side effects from ASM and the need for a driver’s license, and discuss such issues with the patient and their family. As a result, ASM was continued in all four patients in the present study.

Conclusion

Even with 1.5-T ASL, periictal hyperperfusion could be visualized when imaging was performed within approximately 11 h following a short ASS in the present study, indicating its potential utility in diagnosing the pathophysiology of ASS. However, EEG has the limitation of not being performed in a timely manner. Currently, we consider it is most appropriate to first perform MRI with ASL to capture periictal hyperperfusion, followed by the verification of the results using subsequent EEG to accurately diagnose the pathophysiology of ASS.

Acknowledgments

We thank Ryoji Shiraki and colleagues at the Hachisuga Hospital, for supporting our study. We thank Editage for editing the manuscript.

Conflict of Interest Disclosure

The authors declare that they have no conflicts of interest.

References

• [1]  Tanaka T, Ihara M. Post-stroke epilepsy. Neurochem Int 2017; 107: 219–228. PMID:28202284, https://doi.org/10.1016/j.neuint.2017.02.002
• [2]  Haapaniemi E, Strbian D, Rossi C, Putaala J, Sipi T, Mustanoja S, Sairanen T, Curtze S, Satopää J, Roivainen R, Kaste M, Cordonnier C, Tatlisumak T, Meretoja A. The CAVE score for predicting late seizures after intracerebral hemorrhage. Stroke 2014; 45: 1971–1976. PMID:24876089, https://doi.org/10.1161/STROKEAHA.114.004686
• [3]  Galovic M, Döhler N, Erdélyi-Canavese B, Felbecker A, Siebel P, Conrad J, Evers S, Winklehner M, von Oertzen TJ, Haring HP, Serafini A, Gregoraci G, Valente M, Janes F, Gigli GL, Keezer MR, Duncan JS, Sander JW, Koepp MJ, Tettenborn B. Prediction of late seizures after ischaemic stroke with a novel prognostic model (the SeLECT score): a multivariable prediction model development and validation study. Lancet Neurol 2018; 17: 143–152. PMID:29413315, https://doi.org/10.1016/S1474-4422(17)30404-0
• [4]  Sinka L, Abraira L, Imbach LL, Zieglgänsberger D, Santamarina E, Álvarez-Sabín J, Ferreira-Atuesta C, Katan M, Scherrer N, Bicciato G, Terziev R, Simmen C, Schubert KM, Elshahabi A, Baumann CR, Döhler N, Erdélyi-Canavese B, Felbecker A, Siebel P, Winklehner M, von Oertzen TJ, Wagner JN, Gigli GL, Serafini A, Nilo A, Janes F, Merlino G, Valente M, Zafra-Sierra MP, Bayona-Ortiz H, Conrad J, Evers S, Lochner P, Roell F, Brigo F, Bentes C, Peralta AR, Pinho e Melo T, Keezer MR, Duncan JS, Sander JW, Tettenborn B, Koepp MJ, Galovic M, Galovic M. Association of mortality and risk of epilepsy with type of acute symptomatic seizure after ischemic strokes and an updated prognostic model. JAMA Neurol 2023; 80: 605–613. PMID:37036702, https://doi.org/10.1001/jamaneurol.2023.0611
• [5]  Kanazawa Y, Morioka T, Arakawa S, Furuta Y, Nakanishi A, Kitazono T. Nonconvulsive partial status epilepticus mimicking recurrent infarction revealed by diffusion-weighted and arterial spin labeling perfusion magnetic resonance images. J Stroke Cerebrovasc Dis 2015; 24: 731–738. PMID:25724245, https://doi.org/10.1016/j.jstrokecerebrovasdis.2014.09.026
• [6]  Shimogawa T, Morioka T, Sayama T, Haga S, Kanazawa Y, Murao K, Arakawa S, Sakata A, Iihara K. The initial use of arterial spin labeling perfusion and diffusion-weighted magnetic resonance images in the diagnosis of nonconvulsive partial status epileptics. Epilepsy Res 2017; 129: 162–173. PMID:28092848, https://doi.org/10.1016/j.eplepsyres.2016.12.008
• [7]  Oketani H, Morioka T, Inoha S, Miki K, Shimogawa T, Mukae N, Sakata A, Shigeto H, Yoshimoto K. Detection of slightly increased signals in temporal lobe on periictal arterial spin labeling perfusion image for initial suspicion of temporal lobe epilepsy. Epilepsy & Seizure 2023; 15: 67–74. https://doi.org/10.3805/eands.15.67
• [8]  Morioka T, Inoha S, Mugita F, Oketani H, Shimogawa T, Mukae N, Maehara N, Akiyama T, Miki K, Karashima S, Sakata A, Shigeto H, Yoshimoto K. Addition of 1.5-Tesla arterial spin labeling magnetic resonance perfusion imaging to routine electroencephalography in pathophysiological diagnosis of first-onset generalized convulsive seizures in patients with dementia at neurological emergency. Epilepsy & Seizure 2024; 16: 29–43. https://doi.org/10.3805/eands.16.29
• [9]  Morioka T, Mugita F, Inoha S, Akiyama T, Miki K, Shimogawa T, Mukae N, Sakata A, Shigeto H, Yoshimoto K. Usefulness and limitation of addition of 1.5-Tesla arterial spin labeling magnetic resonance perfusion imaging to routine electroencephalography in pathophysiological diagnosis of post-intracerebral hemorrhage epilepsy at neurological emergency. Epilepsy & Seizure 2024; 16: 78–94. https://doi.org/10.3805/eands.16.78
• [10]  Morioka T, Mugita F, Inoha S, Akiyama T, Haruyama H, Karashima S, Shimogawa T, Mukae N, Sakata A, Shigeto H, Yoshimoto K. Ictal 1.5-Tesla MR imaging with addition of arterial spin labeling perfusion images in three patients with electrographic seizures diagnosed with routine electroencephalography based on ACNS Critical Care EEG Terminology 2021. Epilepsy & Seizure 2025; 17: A000162. https://doi.org/10.3805/eands.A000162
• [11]  Miyaji Y, Kawabata Y, Joki H, Seki S, Mori K, Kamide T, Tamase A, Nomura M, Kitamura Y, Tanaka F. Arterial spin-labeling magnetic resonance imaging for diagnosis of early seizure after stroke. J Neurol Sci 2015; 354: 127–128. PMID:25982502, https://doi.org/10.1016/j.jns.2015.04.049
• [12]  Murao K, Morioka T, Shimogawa T, Furuta Y, Haga S, Sakata A, Arihiro S, Arakawa S. Various pathophysiological states of acute symptomatic seizures immediately after ischemic stroke, namely “onset seizures,” shown by complementary use of peri‐ictal magnetic resonance imaging and electroencephalography. Neurol Clin Neurosci 2017; 5: 169–177. https://doi.org/10.1111/ncn3.12160
• [13]  Tada Y, Fujihara T, Shimada K, Yamamoto N, Yamazaki H, Izumi Y, Harada M, Kanematsu Y, Takagi Y. Seizure types associated with negative arterial spin labeling and positive diffusion-weighted imaging on peri-ictal magnetic resonance imaging. J Neurol Sci 2022; 436: 120223. PMID:35279594, https://doi.org/10.1016/j.jns.2022.120223
• [14]  Abe K, Shimogawa T, Mukae N, Ikuta K, Shono T, Tanaka A, Sakata A, Shigeto H, Yoshimoto K, Morioka T. Detection of ictal and periictal hyperperfusion with subtraction of ictal-interictal 1.5-Tesla pulsed arterial spin labeling images co-registered to conventional magnetic resonance images (SIACOM). Surg Neurol Int 2023; 14: 84. PMID:37025532, https://doi.org/10.25259/SNI_723_2022
• [15]  Goto K, Shimogawa T, Mukae N, Shono T, Fujiki F, Tanaka A, Sakata A, Shigeto H, Yoshimoto K, Morioka T. Implications and limitations of magnetic resonance perfusion imaging with 1.5-Tesla pulsed arterial spin labeling in detecting ictal hyperperfusion during non-convulsive status epileptics. Surg Neurol Int 2022; 13: 147. PMID:35509552, https://doi.org/10.25259/SNI_841_2021
• [16]  Oketani H, Morioka T, Inoha S, Shimogawa T, Mukae N, Yasukouchi H, Kinoshita Y, Yoshimoto K. Comparison between 1.5-Tesla pseudocontinuous and pulsed arterial spin labeling in visualizing peri-ictal hyperperfusion: A case report. Epilepsy & Seizure 2023; 15: 17–25. https://doi.org/10.3805/eands.15.17
• [17]  Oketani H, Morioka T, Inoha S, Akiyama T, Shimogawa T, Mukae N, Yoshimoto K. Hemodynamics of ictal hyperperfusion in an epileptic patient with dual pathology revealed by arterial spin labeling perfusion image with triple postlabeling delays. Epilepsy & Seizure 2023; 15: 26–34. https://doi.org/10.3805/eands.15.26
• [18]  Oketani H, Morioka T, Inoha S, Miki K, Shimogawa T, Mukae N, Sakata A, Shigeto H, Yoshimoto K. Arterial spin labeling MR perfusion image in differentiating metabolic encephalopathy from non-convulsive status epilepticus in a patient with generalized periodic discharges with triphasic morphology. Epilepsy & Seizure 2023; 15: 59–66. https://doi.org/10.3805/eands.15.59
• [19]  Oketani H, Morioka T, Inoha S, Mugita F, Mukae N, Shimogawa T, Sakata A, Shigeto H, Yoshimoto K. Periictal hyperperfusion in the perituberal cortex revealed by arterial spin labeling perfusion images in a patient with tuberous sclerosis complex. Epilepsy & Seizure 2023; 15: 75–81. https://doi.org/10.3805/eands.15.75
• [20]  Mugita F, Morioka T, Inoha S, Akiyama T, Shimogawa T, Mukae N, Sakata A, Shigeto H, Yoshimoto K. Crossed cerebellar hyperperfusion during periictal and ictal periods revealed by 1.5-Tesla arterial spin labeling magnetic resonance perfusion images. Epilepsy & Seizure 2023; 15: 95–103. https://doi.org/10.3805/eands.15.95
• [21]  Mugita F, Morioka T, Inoha S, Akiyama T, Maehara N, Shimogawa T, Mukae N, Sakata A, Shigeto H, Yoshimoto T. Absence of periictal hyperperfusion on arterial spin labeling magnetic resonance perfusion images in a patient with posttraumatic epilepsy and underdevelopment of ipsilateral internal carotid and middle cerebral arteries. Epilepsy & Seizure 2024; 16: 1–11. https://doi.org/10.3805/eands.16.1
• [22]  Takahara K, Morioka T, Shimogawa T, Haga S, Kameda K, Arihiro S, Sakata A, Mukae N, Iihara K. Hemodynamic state of periictal hyperperfusion revealed by arterial spin-labeling perfusion MR images with dual postlabeling delay. eNeurologicalSci 2018; 12: 5–18. PMID:30229134, https://doi.org/10.1016/j.ensci.2018.06.001
• [23]  Haga S, Morioka T, Shimogawa T, Akiyama T, Murao K, Kanazawa Y, Sayama T, Arakawa S. Arterial spin labeling perfusion magnetic resonance image with dual postlabeling delay: A correlative study with acetazolamide loading 123I-Iodoanphetamine single-photon emission computed tomography. J Stroke Cerebrovasc Dis 2016; 25: 1–6. PMID:26387043, https://doi.org/10.1016/j.jstrokecerebrovasdis.2015.08.025
• [24]  Haga S, Morioka T, Kameda K, Takahara K, Amano T, Tomohara S, Takaki H, Tsurusaki Y, Arihiro S. Subtraction of arterial spin-labeling magnetic resonance perfusion images acquired at dual post-labeling delay: potential for evaluating cerebral hyperperfusion syndrome following carotid endarterectomy. J Clin Neurosci 2019; 63: 77–83. PMID:30738738, https://doi.org/10.1016/j.jocn.2019.01.044
• [25]  Akiyama T, Morioka T, Shimogawa T, Haga S, Sayama T, Kanazawa Y, Murao K, Arakawa S. Arterial spin-labeling magnetic resonance imaging with dual post-labeling delay in internal carotid artery steno-occlusion: validation with digital subtraction angiography. J Stroke Cerebrovasc Dis 2016; 25: 2099–2108. PMID:27339943, https://doi.org/10.1016/j.jstrokecerebrovasdis.2016.06.005
• [26]  Shimogawa T, Morioka T, Akiyama T, Haga S, Arakawa S, Sayama T. Sequential changes of arterial spin-labeling perfusion MR images with dual postlabeling delay following reconstructive surgery for giant internal carotid artery aneurysm. Surg Neurol Int 2017; 8: 222. PMID:28966828, https://doi.org/10.4103/sni.sni_73_17
• [27]  Tokunaga S, Morioka T, Shirozu N, Tsurusaki Y, Arihiro S, Shimogawa T, Mizoguchi M, Haga S. Arterial spin-labeling perfusion MR images with dual postlabeling delay reveals hemodynamic changes in dural arteriovenous fistulas following endovascular surgery. Interdiscip Neurosurg 2020; 21: 100733. https://doi.org/10.1016/j.inat.2020.100733
• [28]  Shirozu N, Morioka T, Tokunaga S, Shimogawa T, Inoue D, Arihiro S, Sakata A, Mukae N, Haga S, Iihara K. Comparison of pseudocontinuous arterial spin labeling perfusion MR images and time-of-flight MR angiography in the detection of periictal hyperperfusion. eNeurologicalSci 2020; 19: 100233. PMID:32181377, https://doi.org/10.1016/j.ensci.2020.100233
• [29]  Hirsch LJ, Fong MWK, Leitinger M, LaRoche SM, Beniczky S, Abend NS, Lee JW, Wusthoff CJ, Hahn CD, Westover MB, Gerard EE, Herman ST, Haider HA, Osman G, Rodriguez-Ruiz A, Maciel CB, Gilmore EJ, Fernandez A, Rosenthal ES, Claassen J, Husain AM, Yoo JY, So EL, Kaplan PW, Nuwer MR, van Putten M, Sutter R, Drislane FW, Trinka E, Gaspard N. American clinical neurophysiology society’s standardized critical care EEG terminology: 2021 version. J Clin Neurophysiol 2021; 38: 1–29. PMID:33475321, https://doi.org/10.1097/WNP.0000000000000806
• [30]  Szabo K, Poepel A, Pohlmann-Eden B, Hirsch J, Back T, Sedlaczek O, Hennerici M, Gass A. Diffusion-weighted and perfusion MRI demonstrates parenchymal changes in complex partial status epilepticus. Brain 2005; 128: 1369–1376. PMID:15743871, https://doi.org/10.1093/brain/awh454
• [31]  Wieshmann UC, Symms MR, Shorvon SD. Diffusion changes in status epilepticus. Lancet 1997; 350: 493–494. PMID:9274593, https://doi.org/10.1016/S0140-6736(05)63083-8