Clinical Data

70-year-old male patient:

Status post:

• Intracranial hypertensive right basal ganglia bleed a year ago, resulting in mild left-sided hemiparesis.

• Small left cerebellar ischemic infarction half a year ago.

Now presents with worsening of left-sided hemiparesis (left-sided facial droop, severe weakness of the upper/moderate lower limbs) and aphasia. Onset of symptoms two hours ago.

The patient had similar transient neurologic episodes in the preceding week, which improved spontaneously, then presumed to represent transient ischemic attacks (TIAs).

A non-contrast head CT, CT angiogram (CTA) of the cervical and intracranial arteries and CT perfusion (CTP) were performed (code stroke).

What do you see?

• No signs of acute/hyperacute ischaemic infarction in the right (or left) cerebral hemisphere.

• No signs of acute intracranial haemorrhage.

• Stripe of encephalomalacia/gliosis in the region of right basal ganglia, external capsule and corona radiata – location of the old intracranial hemorrhage.

• Volume loss of the crus of right mesencephalon following Wallerian degeneration.

• Old infarction in the left cerebellum.

The perfusion maps show a large area supratentorially on the right with changes of the perfusion parameters, consistent with hypoperfusion (decreased CBV and CBF values, increased TTP and Tmax values).

CBV: cerebral blood volume
CBF: cerebral blood flow
TTP: time to peak
Tmax: time to maximum

Is the patient a candidate for intravenous thrombolysis, mechanical clot retrieval, both, or neither of the two treatment methods?

Neither of the two treatment options, somewhat surprisingly.

Based on the CTP maps, which arterial territory/territories are involved? Which arteries do you think are occluded?

The perfusion abnormalities span the majority of the right cerebral hemisphere, sparing only the region of the right basal ganglia. It seems that the regions of both the anterior and posterior circulation are involved.

Speaking in terms of vascular territories, we would expect such abnormalities if the anterior cerebral artery (ACA), M2 medial cerebral artery (MCA) segment, and the posterior cerebral artery (PCA) on the right were simultaneously occluded.

Internal cerebral artery (ICA) occlusions also cause abnormalities in the ACA and MCA territories, yet an ICA occlusion would also cause abnormalities in the M1 segment of the MCA (the region of the basal ganglia, which is not involved in our case).

The PCA can be supplied by the ICA through a prominent posterior communicating artery, yet in the majority of the cases it gets blood supply mostly from the basilar artery (which is supplied by the vertebral arteries).

So we have an odd situation, where we would be able to explain the perfusion abnormalities by the simultaneous occlusion of an artery of posterior circulation (PCA) and two arteries of the anterior circulation – ACA and part of the MCA. The sparing of the basal ganglia regions argues against proximal ICA occlusion.

Another explanation might be that we are not dealing strictly with a vascular problem.

Is it important to review the CTA image?

Yes, absolutely!

Showing CTA – intracranial arteries, VRT’s (Coronal and oblique/lateral)

CTA: parasagittal MIP’s, right and left carotid bulb

The CTA images demonstrate patent carotid and intracranial arteries (including all of the extra- and intracranial arteries on the right).

What is the diagnosis?

Seizure-related perfusion changes in a patient with postictal paralysis and post-stroke epilepsy.

Stroke is a term used to describe a sudden onset of focal neurological deficit/deficits of presumed vascular origin. It is most commonly caused by arterial occlusion (ischaemic stroke) or an intracerebral haemorrhage (haemorrhagic stroke).

If done promptly enough, arterial occlusions in ischaemic stroke might be resolved by administering i.v. thrombolysis. Occlusions of larger/more proximal arteries might be treated endovascularly by mechanical clot retrieval.

Although stroke is fundamentally a clinical diagnosis, radiology plays an important role in the management of patients, especially in selecting the subgroup who might benefit from the two treatment methods mentioned above.

Stroke is not the only entity causing a sudden onset of focal neurological deficits. Stroke mimics are nonischaemic conditions with clinical presentations similar to stroke. Because of their non-vascular pathophysiology, treating them with i.v. thrombolysis or mechanical clot retrieval is not indicated, as these treatment methods carry the risk of causing intracerebral haemorrhage.

The two of the most common stroke mimics are seizures with postictal paralysis and migraines.

Seizures are transient neurological events characterized by signs/symptoms caused by abnormal increased and synchronous activity of the neurons in the brain. They are common, often single events, and can be provoked by drugs, sleep deprivation or fever. Seizures might cause transient focal neurological deficits after they subside and thus mimic stroke – a term known as postictal/Todd paralysis.

If unprovoked seizures keep recurring in an individual, the term epilepsy is used. Epilepsy might be associated with structural brain lesions of various etiologies, which serve as an epileptogenic focus.

Damaged brain tissue from old ischaemic or haemorrhagic infarcts might also represent an epileptogenic focus and a subset of patients after stroke suffer from post-stroke epilepsy. Re-infarctions and seizures are two main differential diagnoses in post-stroke patients, presenting with worsening of their neurological status.

CT perfusion plays an important role in hyperacute and acute ischaemic stroke imaging. If medial or large vessels are occluded (MeVO and LVO), perfusion parameters are changed in the corresponding brain regions supplied by these arteries, i.e., the perfusion changes in the respective vascular territories. CTP most commonly demonstrates areas of hypoperfusion in the affected regions with increased TTP and TMax values, decreased CBF values and variable CBV values (decreased in ischaemic core and normal/ increased in the ischaemic penumbra).

Stroke mimics might not only cause clinical findings imitating stroke, but also cause CTP changes that can mistakenly be attributed to ischaemic stroke.

CTP changes related to seizures are most often explained by the concept of neurovascular coupling, which states that increased neuronal activity leads to a proportionate increase in blood flow and vice versa. We would expect that imaging a patient at the moment of seizure (ictal phase) would show areas of hyperperfusion in the corresponding part of the seizing brain, owing to the increased neuronal firing. Imaging after the seizure (post-ictal phase) would show signs of hypoperfusion because of the neuronal exhaustion and decreased activity. There is however some overlap between the findings of hypo- and hyperperfusion in clinical practice and CTP cannot be reliably used to differentiate the ictal and post-ictal phases of seizures.

The hypoperfusion pattern is especially important, as the changes in CTP parameters are similar to those observed in stroke. The involved regions in seizure-related changes however do not follow the vascular territories. The described patterns include changes in the whole hemisphere (holohemispheric pattern), lobar, multilobar, and cortical patterns. Correlation with CTA is of utmost importance in these cases, as the examination shows patent arteries in post-seizure-related changes.

Our case revisited

The CTP study showed a holohemispheric pattern of hypoperfusion in the region of the right cerebral hemisphere – changes are not consistent with a vascular distribution. The patency of the vessels was also confirmed by the CTA examination. It was suggested that the CTP findings might be post-seizure related and the neurological deterioration could be a manifestation of postictal paralysis.

The patient did not receive i.v. thrombolysis and also mechanical thrombectomy wasn’t attempted. His neurological symptoms improved spontaneously in the following hours after the CT examination.

After obtaining additional history, it became clear that the patient experienced similar transient neurological episodes in the preceding three months. They were also preceded by twitching of the muscles in his left upper extremity – retrospectively representing seizures.

On the basis of the clinical, imaging and subsequent EEG findings, a diagnosis of post-stroke epilepsy was made. The region of the older haemorrhagic infarction was presumed to represent the epileptogenic focus. The patient was discharged from the hospital on an antiepileptic medication.

References

• Prodi, E., Danieli, L., Manno, C., Pagnamenta, A., Pravatà, E., Roccatagliata, L., Städler, C., Cereda, C., & Cianfoni, A. (2021). Stroke mimics in the acute setting: Role of multimodal CT protocol. American Journal of Neuroradiology, 43(2), 216–222. https://doi.org/10.3174/ajnr.a7379

• Tranvinh, E., Lanzman, B., Provenzale, J., & Wintermark, M. (2018). Imaging evaluation of the adult presenting with New-Onset seizure. American Journal of Roentgenology, 212(1), 15–25. https://doi.org/10.2214/ajr.18.20202

• Phillips, A. A., Chan, F. H., Zheng, M. M. Z., Krassioukov, A. V., & Ainslie, P. N. (2015). Neurovascular coupling in humans: Physiology, methodological advances and clinical implications. Journal of Cerebral Blood Flow & Metabolism, 36(4), 647–664. https://doi.org/10.1177/0271678×15617954

• Van Cauwenberge, M. G., Dekeyzer, S., Nikoubashman, O., Dafotakis, M., & Wiesmann, M. (2018). Can perfusion CT unmask postictal stroke mimics? Neurology, 91(20). https://doi.org/10.1212/wnl.0000000000006501

• Adams, M., & Sharma, R. (2017). Todd paralysis. Radiopaedia.org. https://doi.org/10.53347/rid-57248

• De Cocker, L., D’Arco, F., Demaerel, P., & Smithuis, R. (2012, September 1). Epilepsy – Role of MRI. The Radiology Assistant. Retrieved June 25, 2025, from https://radiologyassistant.nl/neuroradiology/epilepsy/role-of-mri

Clinical Data

2-year-old male:

• Presented with enlarged skull and suspected hydrocephalus

Indicate the abnormality:

Description:

• Mild symmetrical dilatation of the supra & infratentorial ventricular system. Crowded foramen magnum with an inferior location of the cerebellar tonsils below the margins of the foramen magnum.

Which of the following statements are correct?

• Chiari I malformation characterized by a caudal descent of the cerebellar tonsils through the foramen magnum

• Treatment with posterior decompression is usually reserved for symptomatic patients or those with a syrinx

• Chiari I malformations may be seen in association with cervical cord syrinx or hydrocephalus

• Becoming symptomatic is proportional to the degree of descent of the tonsils

• Abnormal skull base , cervical segmentation anomalies, small cranial vault, and posterior fossa could be resembling Chiari I malformation

• Excessive brain tissue could represent acquired Chiari malformation

Come back next week to see the answer. In the meantime, check our social networks to leave your guesses!

Facebook
Instagram
Twitter

Clinical Data

97-year-old polymorbid female:

• Acute onset left hemiplegia (Time of last known well two hours ago)

• Code stroke activated

• Imaging in our institution constituted of:

• Non-contrast head CT

• CTA of aortocervical and intracranial vessels

• CT perfusion

Non-contrast head CT:

CT perfusion revealed a large ischemic core compared with the penumbra (approximately 2:1)

Based on the CTA and non-contrast head CT, which arteries were occluded?

• The CTP shows perfusion changes in the vascular territories of both the right MCA and ACA.

• We know that the MCA is also most certainly occluded because of the dense MCA sign.

• Based on this, there is either a simultaneous occlusion of both the MCA and the ACA or the occlusion is even more proximal – ICA.

CTA shows occlusion of the M1 segment of the right MCA, and the terminal segment of the ICA is patent

However, more proximal intracranial segments of the right ICA were not opacified with contrast

The cervical ICA segments were also not opacified with contrast.(ECA and vertebral artery on the right, no ICA)

Modified sagittal reformation of the bifurcation

Given the CTA appearance of the right ICA, which of the following are the reasonable differentials?

• Occlusion of the right carotid bulb with near total distal ICA occlusion

• ICA dissection

• ICA pseudo-occlusion

Solution:

• Occlusion of the right carotid bulb with near total distal ICA occlusion – FALSE

• ICA dissection – TRUE

• ICA pseudo-occlusion – TRUE

Explanation:

• The carotid bulb is patent and contrast opacification stops 1 to 2 centimeters after the bulb. If the bulb would be really occluded, contrast opacification would stop (almost) at the level of the bulb and the atherosclerotic changes of the bulb would be more extensive.

• The sagittal images nicely show the gradual tapering of contrast opacification in the proximal cervical ICA, which is known as the flame sign.

• The flame sign can be seen in carotid artery dissections and pseudo-occlusions.

How would you differentiate a pseudo-occlusion from a real dissection or ICA occlusion, while the patient is still lying in the CT scanner?

• Obtain a delayed CTA scan of the arteries.

• In other words, repeat the CTA scan at a later time point. Preferably tell the technician at the moment you see the flame sign on the console to just fire the CTA scan again (additional contrast application is not needed).

• (In our case, the clinical context might also have helped. Spontaneous carotid artery dissections are the leading cause of stroke in younger/middle-aged patients. Our patient was almost 100 years old, which made dissection not the most probable diagnosis.)

Normal vs delayed CTA: the unopacified parts of the ACI are now opacified:

Carotid artery pseudo-occlusion:

• A slightly confusing term, as the carotid artery is still occluded, just not in the extent you might think!

• Carotid artery pseudo-occlusion is seen, when there is an occlusion of a distal segment (usually one of the intracranial ones) of the ICA. This occlusion makes it harder for the contrast bolus to flow through the normal ICA – as the cervical ICA segment has no branches. So, using normal timings for the CTA bolus, the cervical segment might still be unopacifed, compared with the normal side. If we wait a little bit and obtain a delayed scan, we give the ICA time to fill up and can assess the extent of the occlusion more accurately. This can be achieved with a delayed CTA scan and is also easily identified with DSA (today mostly performed as the first part of mechanical thrombectomy).

The cavernous segment was unopacified in both the normal and delayed CTA – representing the true level of the occlusion

What are the final diagnoses?

• Right MCA and right cavernous segment ICA occlusion with resultant ischemic stroke

• Pseudo-occlusion of the proximal (mostly cervical) parts of the ICA

Clinical Data

38-year-old female (outpatient setting):

• Appointed for a non-contrast brain MRI

• Requested card information:

• Vertigo

• Unsteady gait

• Headache

• Chronic onset of symptoms that appeared approximately three years ago and got worse in the preceding weeks

T2-weighted axial image:

• The MRI sequences demonstrated no signs of acute or chronic ischaemic lesions, also no mass lesions were detected.

• The signal of the white and grey matter was also otherwise normal.

• Supratentorially, no abnormalities were detected.

T2-weighted axial image:

• Infratentorially, however, the brainstem appeared compressed anteriorly by the dens.

An additional sagittal T2-weighted image was obtained to clarify the anatomy and the cause of obstruction.

• The sagittal slices demonstrated several congenital abnormalities.

Which of the congenital abnormalities contributes the most to the patient’s symptoms?

Which of the congenital abnormalities contributes the most to the patient’s symptoms?

• Basilar invagination (in combination with mild retroflexion of the dens).

• The high riding dens causes compression of the brainstem at the pontomedullary junction.

What other congenital anomaly is basilar invagination commonly associated with?

• Platybasia

Definition of terminology:

Congenital basilar invagination and platybasia are often seen together.

• Basilar invagination:

• A congenital or acquired condition, where the dens protrudes more cranially than it should be, typically above the foramen magnum.

• If the dens causes compression of the brainstem, symptoms might ensue.

• Platybasia:

• A congenital or acquired condition, denoting abnormal flattening of the skull base, formally described as an increased base of skull angle. A simple way to think about it is also a shape of the base of the skull, where the clivus lies more horizontally than it should be (the clivus normally has a downward sloping shape, as clivus is a latin word for slope/hill).

• Translated directly from Greek, platybasia means “flat base” (of skull). And platypus means “flat foot”, just in the case you are wondering.

• Platybasia does not cause symptoms on its own.

The sagittal slices demonstrated several congenital abnormalities:

A CT of the cervical spine and base of skull was recommended to further evaluate the relationship of the bones:

• Basilar impression, platybasia and clivus hypoplasia were redemonstrated.

• CT also showed an additional ossicle/unfused bone immediately superior to the tip of the dens/posterior to the clivus – in the same place as the structure, that appeared to be the sclerotic tip of the dens.

Sagittal and axial reformats:

• The reformats demonstrated a bony arch, which were isolated from the rest of the occipital bone, representing the very rare variant called prebasioccipital arch.

Sagittal and axial reformats, what would we expect in a normal examination from a different patient?

We are still not finished (Part 1):

• Hypoplasia of the occipital condyles, which appear flattened. They are commonly observed together with the aforementioned anomalies.

• Oblique sagittal reformats right and left occipital condyle.

Condylar hypoplasia, oblique coronal reformat.

Oblique right sagittal and coronal reformats, what we would expect (normal examination from a different patient)?

• The normal occipital condyles are convex.

We are still not finished (Part 2):

• Additionally, the angle of the foramen magnum and the plane of the atlas were tilted cranially and posteriorly, resulting in a so-called “lordotic angulation”. The atlas was also noted to be hypoplastic.

Sagittal reformat, what we would expect (normal examination from a different patient)?

Wait a minute! The dens doesn’t actually protrude through the foramen magnum, as the foramen itself is tilted. So…

Are we still dealing with a case of basilar invagination?

Yes, we are.

Although the usual definition of basilar invagination says that the dens should protrude through the foramen magnum, the condition might be defined by several craniometric lineas and measurements. They also take into account conditions, where the foramen magnum itself is abnormally positioned, as in our case.

On sagittal images, they include the McRae, McGregor and Chamberlain lines. On coronal images, the digastric and bimastoid line.

The following Radiopaedia link has them all covered for you: https://radiopaedia.org/cases/basilar-invagination-measurements

In our case, we used the Chamberlain line. It is a line connecting the posterior edge of the bony hard palate to the opisthion (the midline point of the posterior edge of the foramen magnum). If the tip of the dens lies more than 3 mm above this line, we are dealing with basilar invagination.

Chamberlain line; our case and a normal CT of the cervical spine. Basilar invagination is defined, if the tip of the dens lies more than 3 mm above this line (the measurement in our case was 22 mm).

Platybasia measurements

Platybasia is also defined by a craniometric measurement, namely by an abnormally high base of skull angle (>143°).

On CT and MRI, it is defined as angle between the line drawn from the bottom of the anterior cranial fossa (or the nasion) and the tuberculum sellae, and the line drawn from the tuberculum sellae down the posterior margin of the clivus.

Base of skull angle, our case and a normal CT of the cervical spine.

An angle measuring more than 143° defines platybasia. The measurements of the angle were 146° and 67°, respectively.

Conclusion:

When confronted with a congenital anomaly, always be on the lookout, as they are often not isolated.

In the region of the craniocervical junction, congenital basilar invagination and platybasia often occur together.

It is important though to try to put emphasis on the most clinically relevant anomalies. In our case, among the many findings, basilar invagination was the most important one, causing symptoms by exerting pressure on infratentorial structures. Severe cases of basilar invagination can be treated by neurosurgical operative decompression.

Congenital anomalies of the craniocervical junction represent a complex topic. When confronted with difficult or ambiguous cases, looking up and reviewing many craniometric lines and measurements might be helpful.