A systematic approach to detection of epileptogenic foci on imaging: protocols,entities,search pattern,and reporting tools

Abstract

Accurate detection of epileptogenic lesions on imaging requires a systematic search for abnormalities with diverse etiologies, appearances, and locations. This article provides a practical approach to the detection of epileptogenic foci for radiologists. First, we review protocols tailored to seizure evaluation. Next, we present a pictorial review of structural etiologies for seizure, including temporal lobe pathologies, developmental malformations, neoplastic lesions, vascular lesions, and infectious/inflammatory processes. We provide a step-by-step, systematic approach to ensure detection of the manifold potential pathologies visible on imaging as well as appropriate correlation with clinical findings. Finally, we present example reporting templates which can ensure efficient reporting by radiologists as well as serve as mnemonic aids.

Keywords

Epilepsy MRI sequences pictorial review reporting template seizure evaluation systematic approach

Introduction

Epilepsy is a neurological disorder in which a person has a risk of recurrent seizures due to a chronic underlying process.¹ Epilepsy is diagnosed after 2 recurrent unprovoked seizures occurring more than 24 hours apart, after a single unprovoked or reflex seizure with a high risk of reoccurrence (at least 60% risk over the next 10 years), or diagnosis of an epilepsy syndrome.^2,3 The incidence of epilepsy in the U.S. is 79 cases per 100,000 persons with a prevalence of 8 persons per 1000.⁴ Epilepsy exists in many forms with multiple causes encompassing structural, genetic, infectious, metabolic, immune, and unknown etiologies.^5,6 An important aspect of diagnostic evaluation for epileptic seizure is classification based on seizure type: focal, generalized, or unknown.^2,6

Imaging is warranted in most adult patients presenting with a new-onset epileptic seizure, especially focal-onset seizure, or early epilepsy to identify underlying structural etiologies (Table 1).^2,5,7,8 In patients with a known seizure disorder, additional indications for imaging include refractory seizures, changes in seizure pattern, and surgical evaluation.⁹ In the pediatric population, imaging is recommended in children under age 3 presenting with a new-onset and afebrile seizure and in neonates presenting with a seizure.^3,8 Imaging is generally not required in most children with normal neurodevelopment presenting with generalized onset seizure and unambiguous history and examination suggesting a benign seizure disorder, such as simple febrile seizure or absence epilepsy.³

Table 1.

Structural causes of epilepsy grouped by imaging abnormality.

Temporal lobe	Developmental	Epileptogenic tumors	Miscellaneous	Neurocutaneous syndromes
Hippocampal sclerosis (most common cause of focal impaired awareness epilepsy in adults)	Focal cortical dysplasia	Ganglioglioma	Gliosis	Sturge–Weber syndromes
Incomplete hippocampal inversion	Heterotopic gray matter	DNET	Cavernous malformation	Tuberous sclerosis
	Polymicrogyria	Hypothalamic hamartoma	Autoimmune encephalitis
	Polymicrogyria	Pleomorphic astrocytoma	Meningoencephalitis
Uncommon or under-appreciated etiologies
Temporal lobe encephaloceles		MVNT	Schizencephaly
			Rasmussen encephalitis
			Leigh disease

In the nonemergent setting, neuroimaging with MRI using specialized seizure or epilepsy protocols is performed for evaluation of structural brain lesions such as infarction, infection, trauma, tumors, vascular malformation, developmental abnormalities, and other seizure-associated brain pathologies.⁹ In the emergent setting, CT is the initial modality used to rapidly identify immediately treatable structural pathologies while retaining ready access to the patient.^5,9,10 The presence of any of these entities must then be further evaluated with integration of clinical semiology and electroclinical studies by the epileptologist to establish a causal relationship between the structural lesion and the patient’s epilepsy. Identification of an epileptogenic lesion on MRI should prompt referral to a specialized epilepsy surgery center for surgical evaluation as the presence of a structural lesion is a strong indicator of anti-epileptic drug resistance.¹⁰ Furthermore, small lesions identified by MRI and electroclinical studies may be precisely targeted within highly eloquent regions using minimally invasive procedures such as stereotactic radiofrequency thermoablation and newer stereotactic laser ablation.^11,12

Imaging protocol/advanced techniques

Specialized MRI epilepsy protocols are crucial for detection of various structural pathologies that may be missed by routine MRI.^9,13 The ILAE Neuroimaging Task Force identified a set of sequences known as HARNESS-MRI, consisting of high-resolution 3D T1-weighted gradient echo sequence (no interslice gap) with isotopic voxels of 1 mm or less for optimal evaluation of brain anatomy and morphology, high-resolution 3D fluid-attenuated inversion recovery (FLAIR) turbo spin echo sequence (no interslice gap) with isotropic millimetric voxel resolution for assessing signal anomalies, and high in-plane resolution 2D coronal T2WI turbo spin echo sequence with submillimetric voxel resolution (e.g., 0.4 × 0.4 × 2 mm³, no interslice gap) for assessing hippocampal internal structure, amygdala, and parahippocampal cortices.¹⁰ An important caveat for FLAIR is that it is not sensitive for epileptogenic lesions in children before 24 months due to incomplete myelination.¹⁰ Additional MRI sequences include T1WI with gadolinium for evaluation of suspected tumor, vascular malformation, or infectious process and susceptibility-weighted imaging for detection of venous blood, hemorrhage, iron deposits, and calcifications.¹⁰ Several of the entities discussed in this review, such as hippocampal sclerosis and cortical malformations, are better appreciated on scanners with magnetic fields of 3T or higher due to increased signal-to-noise ratio and resolution.¹³

Despite use of these specialized epilepsy protocols, patients with epileptogenic foci may still have visually normal MRI, which has prompted efforts to develop advanced neuroimaging sequences.¹⁴ Double inversion recovery (DIR) is a newer MRI structural protocol with high contrast between gray and white matter for better detection of epileptogenic lesions compared to FLAIR, T1WI, and T2W2 in temporal lobe epilepsy (TLE) and extratemporal focal epilepsy.¹⁴ Fluid and white matter suppression (FLAWS) is a technique that suppresses white and cerebrospinal fluid signals to generate gray matter-specific images for better detection of focal cortical dysplasia. Edge-enhancing gradient echo (EDGE) is another technique for focal cortical dysplasia that directly visualizes the boundary between gray and white matter.¹⁴

Epileptogenic entities

A wide spectrum of etiologies may produce seizures, only a subset of which can be detected via imaging. Non-structural factors contributing to seizures include metabolic derangements, channelopathies, and genetic syndromes. Structural abnormalities identifiable on neuroimaging include temporal lobe pathologies, developmental malformations, neoplastic lesions, vascular lesions, as well as infectious and inflammatory processes. Even when structural abnormalities are present, it is crucial to incorporate electrophysiological information and to pinpoint the abnormal structures that are truly epileptogenic foci.

Temporal lobe etiologies

Hippocampal sclerosis (HS), also known as mesial temporal sclerosis, is the most common cause of medically refractory epilepsy.¹³ The underlying pathophysiology characterized by hippocampal neuronal loss and gliosis. In patients undergoing epilepsy surgery, HS was identified within specimens in 44% of adults and 15% of children.¹⁵ Mesial temporal lobe epilepsy syndrome is associated with focal seizures with impaired consciousness, history of febrile seizures, family history of epilepsy, early onset, and intractable seizures not responsive to anti-epileptic medications. Findings on imaging may be bilateral but are usually unilateral or asymmetric.⁹ Characteristic features include hippocampal volume loss or atrophy, abnormal T2 hyperintensity, and obscuration of internal architecture (Figure 1). Secondary signs include ipsilateral fornix and mammillary body atrophy, as well as enlarged temporal horn and choroidal fissure.¹⁶ Use of 3T, or higher, scanners enables quantitative assessment of hippocampal volume and visualization of hippocampal internal architecture.¹³ Furthermore, use of DIR has been reported to have superior detection of temporal white matter abnormalities compared to FLAIR.¹⁴ Abnormal findings should be integrated with clinical semiology and electrophysiological findings to identify the epileptogenic focus.

Figure 1.

Hippocampal sclerosis: Volumetric brain analysis (a) and coronal FLAIR (b) of a 36-year-old male with new-onset seizures showing asymmetrically reduced volume of the left hippocampus (oval) in comparison to the right with associated increased FLAIR signal compatible with hippocampal sclerosis (circle).

Developmental etiologies

Focal cortical dysplasia

Malformations of cortical development (MCD) are a group of conditions, often with childhood onset, caused by abnormal cell proliferation or apoptosis, abnormal neuronal migration, or abnormal cortical organization.¹³ Focal cortical dysplasia (FCD) is the most common MCD responsible for extratemporal epilepsy. Prevalence of FCD ranges from 5 to 25% depending on collection and imaging technique.¹⁷ FCD is present in 5–10% of epilepsy patients and has a prevalence of up to 50% in children with intractable focal epilepsy.¹³ Despite its prevalence in intractable epilepsy, it is potentially curable by surgery.

FCD is classified into types and subtypes with varying degrees of abnormalites.¹⁸ FCD type I is characterized by mild cortical dyslamination without dysmorphic neurons and presents with absent or subtle MRI findings, such as mild focal cortical thickening, asymmetrical gyral pattern, and blurring of the gray–white matter junction. Type IIa is characterized by cortical lamination with dysmorphic neurons and is more reliably identified with MRI, presenting with cortical thickening, blurred gray–white matter junction, and hyperintense cortical and subcortical signals. Type IIb presents with even more pronounced imaging findings such as cortical thickening, marked blurring of the white–gray matter junction, and a highly specific but not always present transmantle sign or radial band that appears as T2 hyperintense wedge-shaped signal in the subcortical white matter (Figure 2).^18,19 FCD type III is characterized by features of FCD type I or type II in association with other lesions, such as HS (type IIIa), tumors (type IIIb), vascular malformations (type IIIc), or acquired lesions (IIId).

Figure 2.

Focal cortical dysplasia: Axial (a) and coronal (b) FLAIR and T1 (c) MR images of a 20-year-old female with new-onset of seizures show a region of abnormal high FLAIR signal (circle) with associated T1 hypointensity (circle) in the left middle temporal gyrus with associated blurring of the gray/white matter interface, cortical thickening, and a subtle transmantle sign (arrow) suggestive of FCD type IIb. Axial FLAIR (d) MRI from a 32-year-old female shows an incidental region of faint cortical thickening (circle) and FLAIR hyperintensity (arrow) measuring approximately 15 mm in the posterior superior margin of the left middle frontal gyrus, suggestive of FCD type IIa. Stability of the lesion was demonstrated on multiple follow-up exams. Axial (e) and sagittal FLAIR (f) MRI showing asymmetric thickening (oval) of the right anterior superior and middle frontal gyrus with blurring of the gray–white matter junction (circle) and subtle hyperintensity on FLAIR. Color-coded FDG-PET (g) in the same patient shows focally decreased metabolism in the right frontal lobe (arrows), correlating with the region of focal cortical dysplasia noted on MRI.

Due to subtle or absent signs, FCD type I is much less commonly detected compared to type II.¹⁸ Compared to isolated FCD type I and type II, FCD type III generally presents with more subtle or absent dysplastic changes in the cortex adjacent to the principal lesion and lacks the transmantle sign seen in type IIb.²⁰ Diagnosis and identification of epileptogenic foci is improved with careful review of seizure semiology, EEG, and nuclear medicine studies if available. FDG PET is helpful if lesions are not visible or very subtle on MRI and can assist with identification of epileptogenic foci (Figure 2). Furthermore, assessment of the gray–white matter junction may be enhanced by use of FLAWS, which enhances contrast between the gray and white matter, and EDGE, which improves direct visualization of the gray–white matter boundary.¹⁴ In FCD type III, accurate mapping of the entire epileptogenic zone, typically involving the principal lesion and adjacent FCD lesion, with multimodal evaluation is essential to achieving optimal post-surgical seizure control.²¹ Invasive EEG may be used in cases with no identifiable epileptogenic foci on imaging.¹³ Important alternative diagnoses to consider during evaluation of FCD include tuberous sclerosis, cortical neoplasm (DNET, ganglioglioma, low-grade glioma), incomplete myelination, and gliosis. FCD is distinguished from these other diagnoses by its characteristic focal cortical findings on imaging and lack of systemic clinical features.

Polymicrogyria

Polymicrogyria is a common MCD characterized by small and disorganized gyral convolutions with shallow sulci (“nodular cortex”) resulting from abnormal distribution of cortical neurons. The bilateral perisylvian region is the most commonly affected, but findings may also be unilateral.¹³ It is also often associated with other malformative lesions such as cerebellar hypoplasia, callosum agenesis, periventricular nodular heterotopia, and schizencephaly.⁹ On MRI of patients with complete myelination (greater than 2 years old), findings include an abnormally thickened and irregular cortex instead of small gyri (Figure 3).²² In patients 1–2 years old, polymicrogyria presents with an abnormal sulcation pattern. In neonates (or younger than 1 year old), polymicrogyria is best visualized with thin T2WI. Patients 1–2 years old may also be imaged with 3D spoiled-gradient recalled echo (SPGR) T1WI. Alternative diagnoses to consider include microcephaly with simplified gyral pattern, hemimegaloencephaly, pachygyria, and cobblestone malformations. Polymicrogyria can be distinguished from these other diagnoses by its small, irregular gyri without reduced brain volume, hemispheric enlargement, broad and thick gyri, or disorganized cortex with indistinct gray–white matter junction.

Figure 3.

Polymicrogyria: An 18-month-old with a history of mild global developmental delay, ADHD, buccolingual apraxia and seizures. Axial (a), coronal (b), sagittal right, (c) and sagittal left (d) T1 MR images show thickening and irregularity (ovals) of the perisylvian gray matter bilaterally extending to the perirolandic region, compatible with polymicrogyria.

Schizencephaly

Schizencephaly is another cortical malformation characterized by a cleft extending from the cortical surface to the ventricular ependyma.⁹ It is a rare condition with a prevalence of 1.54 per 100,000 births.²³ The cleft is lined by dysplastic gray matter, often polymicrogyria, and may occur bilaterally or unilaterally, which is often associated with contralateral polymicrogyria. When the walls of the cleft are separated with CSF present in between, often bilaterally, the malformation is termed open-lipped (Figure 4).²⁴ When the walls of the cleft are in apposition without CSF in between, usually in unilateral schizencephaly, the type is close-lipped. Multiplanar sequences with or without volumetric acquisition are particularly useful for close-lipped schizencephaly. In patients with incomplete myelination, the gray matter lined cleft is best seen on T2WI.²⁵ It is important to distinguish schizencephaly from porencephaly cyst, which is an acquired lesion characterized by smooth walls without a gray matter lining or associated cortical malformations, such as polymicrogyria with schizencephaly.²⁶

Figure 4.

Schizencephaly: Axial FLAIR (a), axial T2 (b) show a case of bilateral open-lip (arrows) schizencephaly in a patient with partial agenesis of the corpus callosum.

Heterotopic gray matter

Heterotopic gray matter is a rare disorder of abnormal neuronal migration that results in the formation of nodules of ectopic gray matter, consisting of mature neurons and glial cells, without defined lamellar organization.⁹ Heterotopia may occur anywhere from the periventricular region to the pia. Periventricular (subependymal) nodular heterotopia is the most common subtype (Figure 5).²⁷ Subcortical laminar heterotopia (double cortex) is a subtype that involves a continuous or semicontinuous band of gray matter below the cortical mantle.²⁸ Other subtypes include subcortical nodular heterotopia and pial heterotopias (“cobblestone” cortex). Cobblestone heterotopia is characterized by an irregular, bumpy “cobblestone” brain surface with cortical dysplasia, disruption of the normal cortical contour, and agyria.²⁹ Helpful MRI sequences for detection of heterotopia include 3D T1WI SPGR in patients with complete myelination and 3D T2WI fast spin echo (FSE). Alternative diagnoses to consider include TSC, closed-lip schizencephaly, ependymal spread of a tumor, polymicrogyria, and congenital CMV. Heterotopic gray matter is distinguished by its non-enhancing, isointense gray matter nodules without mass effect or systemic features.

Figure 5.

Heterotopic gray matter: Axial T1 Gd+ (a), and axial FLAIR (b) MRI showing incidental heterotopic gray matter (arrows) in a post-operative study.

Ulegyria

Ulegyria is a condition that usually results from perinatal global hypoxic ischemic brain injury producing cortical and subcortical lesions such as shrunken and flattened cortex and multiple cerebral convolutions (Figure 6).³⁰ In children with a history of term hypoxic-ischemic injury and cortical injury on imaging, ulegyria was identified in 66% of patients.³¹ Ulegyria is most commonly centered around the deepest portion of a gyrus and usually in the parasagittal watershed and posterior cerebral artery distributions.³² On MRI, ulegyria is characterized by a unique mushroom-shaped appearance resulting from small atrophic circumvolutions at the bottom of a sulcus underlying an intact gyral apex. Findings may be unilateral or bilateral. Associated clinical findings include cerebral palsy, mental disability, and intractable focal seizures often with visual auras and mean onset at 4 years old.¹³

Figure 6.

Ulegyria: Axial FLAIR (a) and T1 Gd+ (b) show a shrunken and flattened gyrus (ovals) in the left parietal lobe compatible with ulegyria.

Ectopic posterior pituitary

Ectopic posterior pituitary (EPP) is a rare condition characterized by the absence, truncation, or thread-like appearance of the pituitary stalk leading to positioning of the posterior pituitary gland outside of the sella turcica. EPP is thought to arise from defective neuronal migration during embryogenesis.³³ EPP is best seen on midline sagittal T1WI MRI and is commonly located along the median eminence of the tuber cinereum or truncated pituitary stalk (Figure 7).³⁴ As age increases, the pituitary bright spot dims and may appear absent on T1WI.³⁵ Commonly associated abnormalities include heterotopia, optic nerve hypoplasia, and corpus callosum abnormalities. Alternative diagnoses to consider include surgical or traumatic stalk transection, central diabetes insipidus, and hypothalamic lipoma (in the tuber cinereum). EPP can be distinguished from these diagnoses by gradual progression of panhypopituitarism, presence of ectopic bright spot seen on fat-specific imaging.

Figure 7.

Ectopic posterior pituitary: Sagittal T1 Gd+ (a), sagittal FLAIR (b), and coronal T1 w/o contrast (c) MRI show an intrinsically T1 hyperintense nodule (arrows) in the superior aspect of the pituitary infundibulum which enhances in the early dynamic phase and remains hyperintense on Gd + FS. The pituitary T1 posterior bright spot (arrow) is missing. Findings are suggestive of an ectopic posterior pituitary.

Septo-optic dysplasia

Septo-optic dysplasia (SOD) is a rare congenital neurodevelopmental disorder with a wide range of clinical presentations ranging from normal cognitive and neurological function to developmental delay and epilepsy. SOD has prevalence at birth of 1.41 per 10,000 live births.³⁶ It defined by the presence of at least 2 of the following three features: optic nerve hypoplasia (often asymmetric), midline brain abnormalities (typically malformations in the septum pellucidum and corpus callosum), and pituitary gland dysfunction (often ectopic posterior pituitary).³⁷ There may be flat-roofed ventricles and inferiorly pointed frontal horns (Figure 8).³⁸ The presence of additional cerebral cortical malformations such as polymicrogyria and schizencephaly is known as SOD-plus syndrome. Imaging of SOD involves MRI with 3D acquisitions or multiplanar reformation (MPR) such as 3D GRE-based T1 and 3D-balanced steady state (SS) GRE.

Figure 8.

Septo-optic dysplasia: Coronal T1 (a), coronal T2 (b) MR show a hypoplastic left optic nerve (arrow) with increased T2 signal (arrow). Coronal T2 (c) shows an absent septum pellucidum (arrow) and flat-roofed lateral ventricles (arrow).

Cephalocele

A cephalocele is another congenital anomaly that is characterized by extracranial extension of intracranial contents through a defect in the skull. These contents can include meninges, brain tissue, and CSF. It occurs in 0.8–4 per 10,000 live births.³⁹ Cephalocele more commonly occurs in the occipital region but can also occur at the frontoethmoidal, parietal (atretic or cranial vault), or basal (nasopharyngeal) regions.⁴⁰ On MRI imaging, a cephalocele will have a heterogeneous signal intensity of contents reflecting the brain parenchyma and CSF (Figure 9). MRV can help to characterize venous relationships to the cephalocele, particularly in the occipital region. Alternative diagnoses to consider include dermoid or epidermoid cyst, nasal glial heterotopia (“nasal glioma”), sinus pericranii, and vascular anomalies such as neoplasms and malformations. These alternative diagnoses lack direction connection to the intracranial compartment and a skull defect or will show venous communication, such as in the case of sinus pericranii.

Figure 9.

Cephalocele: Coronal FLAIR (a, b), and axial T2 (c) of the brain showing a right-sided temporal lobe encephalocele (arrows).

Tuberous sclerosis complex

Tuberous sclerosis complex (TSC) is a rare neurocutaneous disorder characterized by hamartomatous lesions or “tubers” in various organs and systems.^9,13 The incidence is estimated to be between 1:6,000 and 1:10,000 live births.⁴¹ TSC often initially manifests as infantile spasms in the first months of life, progressing to intellectual disability and intractable focal seizures. In the CNS, TSC is characterized by cortical tubers, subependymal nodules, subependymal giant cell astrocytoma, and white matter lesions. On MRI imaging, tubers change with myelination, appearing hyperintense on T1 and hypointense on T2 in neonates and appearing hypointense on T1 and hyperintense on T2 with poorly defined borders in older children (Figure 10).^9,42 Cortical tubers may be difficult to differentiate between FCD. Unlike FCD, cortical tubers tend to be multifocal, affect any part of the cortex, are often associated with calcifications, and are present with other features of TSC. It is also important to consider cavernomas as a differential diagnosis in the evaluation of calcified tubers. Other alternative diagnoses include gray matter heterotopia, cortical dysplasias, and germinal matrix. TSC can be distinguished by the presence of subependymal nodules and systemic features.

Figure 10.

Tuberous sclerosis: Axial FLAIR (a-c) and axial SWI (d) show scattered foci of FLAIR hyperintensities (circle) in multiple lobes, some of which show blooming artifact (circle) on SWI, compatible with calcified cortical tubers in the setting of TSC. Subependymal nodules (arrow) and multiple scattered linear or radial bands of FLAIR hyperintensities (arrow) are also shown. Axial FLAIR (e), coronal FLAIR (f), and coronal T2 at multiple planes (g–i) MRI show multiple subtle juxtacortical (circles)/cortical T2/FLAIR hyperintensities (arrows) predominantly in the bilateral frontal lobes, which may represent focal cortical dysplasia versus cortical tubers.

Sturge–Weber syndrome

Sturge–Weber syndrome (SWS) is another rare congenital neurocutaneous disorder characterized by abnormal cortical venous development with an incidence of 3.08 per 100,000 people per year.^43–45 Clinically, patients with SWS often present with focal antiepileptic-resistant seizures, hemiparesis, cognitive impairment, and headaches. The hallmark features of SWS include facial capillary malformation, known as a port-wine stain typically involving the first division of the trigeminal nerve, leptomeningeal vascular malformations, and ocular abnormalities such as glaucoma. Imaging findings include ipsilateral leptomeningeal enhancement, ipsilateral enlarged choroid plexus, cortical atrophy, cortical calcification with a tram track appearance, calvarial thickening, and sinus hyperpneumatization (Figure 11).⁴⁴ Alternative diagnoses to consider include an acquired meningeal process, PHACES syndrome, blue rubber bleb nevus syndrome, Klippel–Trenaunay syndrome, and Wyburn–Mason syndrome. SWS is distinguished from these diagnoses by the presence of a port-wine stain with ipsilateral leptomeningeal enhancement and ocular involvement.

Figure 11.

Sturge–Weber syndrome: Coronal T2 (a), axial T1 (b), and axial FLAIR (c) showing cortical atrophy (circle, arrows) associated with increased signal on FLAIR of the left cerebral hemisphere (arrow), most pronounced in the left frontal lobe, compatible with the patient’s known Sturge–Weber syndrome.

Leigh disease

Leigh disease or subacute necrotizing encephalomyelopathy is a rare neurometabolic disorder characterized by progressive neurodegeneration typically manifesting in infancy or early childhood.⁴⁶ The prevalence is estimated to be between 1:30,000 and 1:40,000.⁴⁷ Clinically, Leigh disease presents with developmental delay, hypotonia, seizures, ataxia, and respiratory. Lactate may be elevated in the blood, CSF, as well as within the brain (on MR spectroscopy) due to mitochondrial dysfunction. The hallmark feature of Leigh disease is symmetrical “speckled” lesions in the basal ganglia or brainstem on MRI, which appear as hyperintensities on T2 and hypointensities on FLAIR (Figure 12).⁴⁸ Alternative diagnoses to consider include hypoxic ischemic encephalopathy, Wilson disease, near drowning, kernicterus, Huntington disease, and encephalitis. Leigh disease is distinguished from these conditions by its progressive neurological symptoms, symmetric lesions in the basal ganglia on MRI, and elevated lactate on various testing modalities.

Figure 12.

Leigh disease: Axial FLAIR (a), T2 (b), and SWI (c) showing increased signal on T2, decreased signal on FLAIR, and susceptibility on SWI (ovals) in the bilateral putamen, compatible with the history of Leigh syndrome in this patient.

Inflammatory/infectious etiologies

Rasmussen encephalitis

Rasmussen encephalitis is a rare immune-mediated progressive hemispheric encephalopathy characterized by childhood onset, intractable epilepsy, hemiparesis, and neurologic decline.⁹ The incidence is estimated to be 2.4 per 10 million people aged less than 18 per year.⁴⁹ On MRI, there is progressive atrophy of one cerebral hemisphere, usually beginning in the insula, perisylvian region, caudate head, and putamen, with cortical hyperintensities on T2 and FLAIR (Figure 13).¹³ Serial imaging may show fluctuations with regression followed by reappearance of lesions on FLAIR in the cortical and subcortical regions, predominantly in the frontoinsular and parietal regions.⁵⁰ There may also be a crossed cerebellar diaschisis. Alternative diagnoses to consider include SWS, other autoimmune encephalitis (often with bilateral findings), and Dyke–Davidoff–Masson syndrome. Rasmussen encephalitis is distinguished by its progressive, unilateral cerebral atrophy with T2/FLAIR hyperintensities and without enhancement.

Figure 13.

Rasmussen encephalitis: Axial T1 (a), coronal T2 (b), and coronal FLAIR (c) MRI of the brain showing left cerebral hemisphere volume loss (arrows, circles) compared to the right, most severe in the parietal lobe, and least severe in the frontal lobe. The caudate head and basal ganglia appear unremarkable.

Meningoencephalitis

Meningoencephalitis is a condition of simultaneous inflammation of the meninges and brain parenchyma caused by a wide variety of infectious and non-infectious agents.⁵¹ It is typically a clinical diagnosis that presents with fever, neck stiffness, altered mental status, seizures, and focal neurological deficits with confirmation by CSF analysis showing pleocytosis, elevated protein, and elevated inflammatory markers. Imaging plays a role in identifying etiologies and assessing complications such as empyema, ischemia, hydrocephalus, cerebritis, and ventriculitis. Location of abnormalities provides a clue for the etiology with cerebral convexities suggesting a bacterial agent; basilar meninges suggesting TB, neurosyphilis, cryptococcus, sarcoidosis, and lymphoma; and meningoencephalitis suggesting a viral agent.⁵² These lesions may appear as hyperintense signals in sulci, cisterns, or brain parenchyma on T2 and FLAIR with enhancement of the meninges (Figure 14). It is important to note that meningitis may be present even with normal imaging.

Figure 14.

Meningoencephalitis: Coronal T1 Gd+ (a), coronal T2 (b), and axial T1 Gd+ (c) showing frontotemporal and subtle cortical enhancement (arrows, circle) with associated increased T2 signal (circles) concerning for meningoencephalitis.

Autoimmune encephalitis

Autoimmune encephalitis (AE) is an immune-mediated disease with antibody-mediated inflammation of the brain. It has an incidence of 0.8 per 100,000 person-years.⁵³ Paraneoplastic AE may occur as a neurologic complication of underlying malignancy outside the CNS, most commonly small cell lung carcinoma.⁵⁴ The most common clinical paraneoplastic syndrome is limbic encephalitis. Clinically, limbic encephalitis presents with rapid onset of psychiatric symptoms, cognitive impairment, temporal lobe seizures, and movement disorders. On imaging, limbic encephalitis is characterized by hyperintensities in the mesial temporal lobes and limbic system (Figure 15).⁵⁵ It is important to note that brain MRI is positive in only 30–50% of patients with anti-NMDAR paraneoplastic encephalitis.⁵⁶ Alternative diagnoses to consider include HSV encephalitis, diffuse astrocytoma, status epilepticus, and gliomatosis cerebri. Limbic encephalitis, which may mimic HSV encephalitis, is typically associated with a subacute or chronic course.

Figure 15.

Paraneoplastic encephalitis: Axial T2 pelvic MR (a) from a patient presenting with multiple episodes of seizures. The patient underwent several brain MRI scans which had failed to reveal an etiology for the seizures. A full-body MRI subsequently revealed an ovarian teratoma (arrow). The seizures were then found to be related to anti-NMDAR encephalitis. Axial FLAIR MR (b) from a different patient shows abnormal hyperintensity (circles) in the bilateral medial temporal lobes, characteristic of limbic encephalitis, the most common paraneoplastic syndrome.

Neoplastic etiologies

Long-term epilepsy associated tumors

Long-term epilepsy associated tumors (LEATs) are encountered in approximately 30% of patients with long-term pharmaco-resistant epilepsy.¹³ The most common of these are gangliogliomas, which appear as cortically based enhancing cystic nodules on imaging (Figure 16).⁵⁷ They are most frequently found in the temporal lobe and can be supratentorial but can also be infratentorial. Calcifications are found in about 50% of cases. Most cases are also associated with cortical disorganization (FCD type IIIb). Differentiating features of FCD include the presence of cortical thickening and transmantle sign. Other alternative diagnoses to consider include pilocytic astrocytoma, DNET, xanthoastrocytoma, diffuse astrocytoma, and oligodendroglioma.

Figure 16.

Ganglioglioma: A 19-year-old female presenting due to “flutter episodes” for approximately a year and one generalized tonic-clonic seizure. Coronal T2 (a), axial FLAIR (b), and axial post-contrast (c) MR images show an approximately 1 cm rounded enhancing intra-axial lesion (circles) in the lateral aspect of the right superior frontal gyrus with no surrounding edema or mass effect. The patient underwent resection of the lesion, and surgical pathology confirmed the diagnosis of ganglioglioma.

Dysembryoplastic neuroepithelial tumor

Dysembryoplastic neuroepithelial tumor (DNET) is a benign, slow growing, mixed-glial neuronal neoplasm.¹³ On imaging, DNETs appear as a T2-hyperintense “bubbly” cortical mass with minimal or no mass effect, no peritumoral edema, and typically with sharp demarcation (Figure 17).⁵⁸ Enhancement is significantly less common than with gangliogliomas. Some DNETs with less distinct margins may have associated focal cortical dysplasia. The most common site is the medial temporal lobe. Alternative diagnoses to consider include ganglioglioma, FCD, neuroepithelial cyst, pleomorphic xanthoastrocytoma, and angiocentric glioma.

Figure 17.

Dysembryoplastic neuroepithelial tumor: A 46-year-old male with new-onset seizures. Coronal T1 (a), T2 (b), FLAIR (c), and axial T2 (d) show a multilobar, T1 hypointense, T2/FLAIR hyperintense, “bubbly” lesion (ovals) in the right temporal lobe. The lesion was resected and pathology was consistent with dysembryoplastic neuroepithelial tumor (DNET).

Multinodular and vacuolating nodular tumor

Multinodular and vacuolating nodular tumor (MVNT) is a rare, benign neuronal tumor characterized by multiple nodules and vacuolated neuronal cells, primarily encountered as an epileptogenic lesion in the subcortical white matter.⁵⁹ MVNT appears as a cluster of variably sized, confluent, and discrete nodules, often in a U-shaped configuration with no mass effect (Figure 18). They can be found in the deep cortical ribbon and superficial white matter sparing the superficial gray matter and can also be found in the vermis of the cerebellum. On MRI, the lesions are isointense with the cortex on T1WI, hyperintense on T2WI/FLAIR, very bright on DWI, and exhibit no enhancement. Alternative diagnoses to consider include DNETs, FCD, and enlarged perivascular spaces.

Figure 18.

Multinodular and vacuolating neuronal tumor: Coronal T2 (a) and FLAIR (b) MRI show confluent and multiple discrete T2/FLAIR hyperintense nodules (ovals) along the undersurface of the cortex within the left inferior parietal lobe. Please note the U-shape configuration of the clusters (arrows) and the lack of mass effect in the magnified coronal T2 (c). Coronal pre- (d) and post-Gd+ (e) T1 MRI show the isointense nodules (circles) compared to the cortex.

Miscellaneous etiologies

Cerebral cavernous malformations

Cerebral cavernous malformations, also known as cavernomas, are benign vascular lesions made up of dilated sinusoids lined by thin, immature walls containing blood products of various ages.⁹ This results in hamartomatous or “berry-like” of vascular spaces without any intervening brain parenchyma.⁶⁰ The peak incidence for cavernomas is 4–60 years old, and they are estimated to occur in 0.5% of the general population.⁶¹ On imaging, cavernomas have a heterogenous or “popcorn” appearance on T2, a hypointense hemosiderin rim, hypointensity on SWI, and absent or minimal enhancement (except in cavernomas with associated developmental venous anomaly) (Figure 19).⁶² Sizes may range from microscopic to giant (larger than 6 cm), and they are most common in the supratentorial region compared to the infratentorial region and spinal cord. Alternative diagnoses to consider include diffuse axonal injury, intracerebral hematoma, neoplasm, and capillary telangiectasia. Cavernomas are distinguished by their “popcorn” appearance, hemosiderin rim, and lack of significant enhancement and edema.

Figure 19.

Cavernous malformation: Axial T2 (a), T1 (b), FLAIR (c), and SWI (d) MRI of the brain showing heterogenous signal (ovals) with a hypointense rim on T2, T1, and FLAIR of the right uncus as well as a decreased signal on SWI.

Gliosis

Gliosis is a relatively common reactive process occurring after some time following most types of CNS injuries, such as trauma, infection, ischemia, and neurodegenerative disease.¹ The prevalence of gliosis varies based on the underlying condition. In patients undergoing epilepsy surgery, glial scars were found in 4.9% of patients.¹⁵ Gliosis is characterized by focal proliferation and activation of glial cells, particularly astrocytes. Gliosis can be asymptomatic or be a seizure-focus. The presenting features of gliosis are typically related to the underlying condition causing the gliosis. On imaging, gliosis produces no volume loss, no mass effect, and appears hyperintense on T2/FLAIR and mildly hypointense on T1, with mildly facilitated diffusion on ADC (Figure 20).⁶³ It may also mimic a low-grade astrocytoma. On CT, small or subtle areas of gliosis may be invisible but more pronounced gliosis produces white matter hypoattenuation.

Figure 20.

Gliosis: Axial T1 (a), T2 (b), and FLAIR (c) MRI of the brain showing decreased signal on T1 and increased signal on T2 and FLAIR associated with bifrontal gliosis (arrows).

Dual pathology

Dual pathology refers to the coexistence of hippocampal sclerosis and another extrahippocampal epileptogenic entity, most commonly MCDs or acquired childhood lesions.¹³ Satisfaction of search after detection of either HS or another epileptogenic entity can result in failure to detect both conditions, ultimately reducing success of epilepsy surgery and highlighting the importance of a systematic approach.¹³

Multimodality imaging

Several of the entities discussed above have subtle findings that make detection difficult, even resulting in labeling as “normal.” These conditions include focal cortical dysplasias, small gangliogliomas, and DNETs.¹³ Detection of these entities can be improved by selection of proper MRI techniques, incorporation of EEG and clinical findings, multimodality imaging with PET and SPECT, and a systemic approach for evaluating the epileptogenic lesions. Chart review should include age of seizure onset, seizure semiology, family history, and suspected localization from EEG findings. Multimodality imaging can also provide additional utility for epilepsy evaluation through confirmation or further mapping of the epileptogenic zone. In addition to being essential in the emergent setting for evaluation of acute hemorrhage, CT can also help detect conditions associated with calcifications. PET is helpful for detection of subtle FCD and surgical planning. SPECT can be used for interictal subtraction for localization, and fMRI is also commonly used for surgical planning. It is essential to avoid overinterpretation of structural lesions as epileptogenic foci without clinical correlation.

Systematic approach

Using a checklist approach or including pertinent positives and negatives in a reporting template can ensure a thorough search for possible etiologies. Grouping pathologies by location and broader categories of imaging appearance can reinforce a comprehensive search for the large diversity of the potential abnormalities, without requiring call-out of every entity. Sites or groupings of pathology that may be useful to specifically enumerate in a search pattern or report template include cortical sulcation abnormalities, hippocampal abnormalities, parenchymal signal abnormalities, subependymal/ventricular lining contour abnormalities, and midline fusion/developmental abnormalities (Table 2).

Table 2.

Epileptogenic entities associated with certain imaging findings on MRI.

Imaging findings	Epileptogenic entities
Abnormal cortical sulcation or contour	FCD, polymicrogyria, ulegyria, microcephaly with simplified gyral pattern, hemimegalencephaly, pachygyria, cobblestone malformations, other MCDs, SWS, Rasmussen encephalitis, encephaloceles
Asymmetry of the hippocampi	Hippocampal sclerosis, IHI
Parenchymal masses or nodules	Epileptogenic tumors (e.g., ganglioglioma, DNET, and metastases), gray matter heterotopia, cavernoma
Abnormal brain susceptibility	Neurocysticercosis, calcified cortical tubers of TSC, cavernomas, mineralization or blood products within other brain lesions
Other abnormal brain signal	Gliosis, ischemia, trauma, infection, metabolic, demyelinating, and autoimmune processes
Abnormal ventricular contour (including subependymal nodularity)	Subependymal nodules of TSC, subependymal/periventricular nodular heterotopia, schizencephaly, ependymal spread of neoplasm
Midline fusion and developmental abnormalities	Holoprosencephaly, agenesis of the corpus callosum, SOD, hypothalamic hamartoma, EPP

For example, detection of abnormal cortical sulcation or contour can suggest entities such as FCD, polymicrogyria, and ulegyria. Hippocampal abnormalities or asymmetries should prompt consideration of hippocampal sclerosis. Detection of a parenchymal mass or nodule can raise suspicion for etiologies such as gray matter heterotopia, ganglioglioma, and other neoplasia. The presence of abnormal brain susceptibility should include consideration of cavernomas, neurocysticercosis, or mineralization or blood products within other cerebral lesions. Other abnormal (e.g., T2/FLAIR) brain signal is associated with a wide range of conditions including gliosis, ischemia, trauma, infection, metabolic, demyelinating, and autoimmune processes. Subependymal/ventricular contour abnormalities can be caused by TSC, schizencephaly, heterotopia, and neoplasms. Conditions with midline fusion abnormalities include holoprosencephaly, agenesis of corpus callosum, and SOD. Entities with other midline developmental abnormalities include EPP and hypothalamic hamartoma. Chunking of etiologies in this way facilitates not just workflows and reporting (Table 3) but also the recollection of epileptogenic lesions detectable on imaging.

Table 3.

Example reporting templates for pertinent positives and negatives in seizure imaging with differing styles.

Example reporting macro with a “structured” format	Cortical sulcation and contour abnormalities: [Absent/Present]
	Parenchymal masses or nodules: [Absent/Present]
	Asymmetry of the hippocampus: [Absent/Present]
	Abnormal brain susceptibility: [Absent/Present]
	Other abnormal brain signal: [Absent/Present]
	Abnormal ventricular contour: [Absent/Present]
	Midline abnormalities: [Absent/Present]
Example reporting macro which enumerates specific pertinent negative categories	No abnormal appearing sulcation pattern
	No parenchymal mass or nodule
	No focal asymmetry of the hippocampi
	No abnormal brain susceptibility
	No other parenchymal signal abnormality
	No ventricular contour abnormality
	No midline fusion abnormality
Example reporting macro which enumerates specific pertinent negatives, with a combination of specific entities and categories	No abnormal appearing sulcation pattern
	No evidence of polymicrogyria
	No focal asymmetry of the hippocampi
	No subependymal nodularity or heterotopic gray matter
	No midline fusion abnormality
	No evidence of schizencephaly
	No encephalocele seen
	No evidence of cavernous malformation
	No ectopic pituitary bright spot
	No mass in the region of the tuber cinereum
Example prose-style macro which enumerates specific common etiologies and categories of imaging abnormality	There is a normal sulcation pattern. There is no polymicrogyria. The hippocampi are normal in signal and morphology bilaterally. There is no subependymal nodularity, heterotopic gray matter, or cortical or subcortical signal abnormality. The region of the tuber cinereum is unremarkable. There is no encephalocele. There is no cavernous malformation

Conclusion

Identification of epileptogenic foci can be facilitated with use of a systematic approach. Consideration of clinical semiology and electrophysiological information, recognition of the features of notable intracranial pathologies, use of specialized epilepsy protocols for certain relatively common conditions such as FCD and hippocampal sclerosis, and use of comprehensive epilepsy reporting templates can facilitate evaluation of structural causes of epilepsy on imaging.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Richard J. Lozano

Carys Kenny-Howel

References

Rao

Lowenstein

, et al. Seizures and epilepsy. In: Loscalzo

Fauci

Kasper

(eds). Harrison's Principles of Internal Medicine. 21e. McGraw-Hill Education, 2022.

Smith

PEM

. Initial management of seizure in adults. N Engl J Med 2021; 385: 251–263.

Fine

Wirrell

. Seizures in children. Pediatr Rev 2020; 41: 321–347.

Helmers

Thurman

Durgin

, et al. Descriptive epidemiology of epilepsy in the U.S. population: a different approach. Epilepsia 2015; 56: 942–948.

Lee

Burns

Ajam

, et al.ACR appropriateness criteria® seizures and epilepsy. Journal of the American College of Radiology. 2020 May 1;17(5):S293-304.

Scheffer

Berkovic

Capovilla

, et al. ILAE classification of the epilepsies: position paper of the ILAE commission for classification and terminology. Epilepsia 2017; 58: 512–521.

Tranvinh

Lanzman

Provenzale

, et al. Imaging evaluation of the adult presenting with new-onset seizure. Am J Roentgenol 2019; 212: 15–25.

Expert Panel on Pediatric Imaging Trofimova

Milla

Ryan

, et al.ACR Appropriateness Criteria® Seizures-Child. J Am Coll Radiol. May 2021;18(5S):S199-S211. doi: 10.1016/j.jacr.2021.02.020.

Cendes

Theodore

Brinkmann

, et al. Neuroimaging of epilepsy. Elsevier, 2016, pp. 985–1014.

10.

Bernasconi

Cendes

Theodore

, et al. Recommendations for the use of structural magnetic resonance imaging in the care of patients with epilepsy: a consensus report from the International league against epilepsy neuroimaging task force. Epilepsia 2019; 60: 1054–1068.

11.

Voges

Büntjen

Schmitt

. Radiofrequency-thermoablation: general principle, historical overview and modern applications for epilepsy. Epilepsy Res 2018; 142: 113–116.

12.

Curry

Gowda

McNichols

, et al. MR-guided stereotactic laser ablation of epileptogenic foci in children. Epilepsy Behav 2012; 24: 408–414.

13.

Adin

Durand

Zucconi

, et al. The changing landscape in epilepsy imaging: unmasking subtle and unique entities. Diagn Interv Radiol 2022; 28: 503–515.

14.

Sone

. Making the invisible visible: advanced neuroimaging techniques in focal epilepsy. Front Neurosci 2021; 15: 699176.

15.

Blumcke

Spreafico

Haaker

, et al. Histopathological findings in brain tissue obtained during epilepsy surgery. N Engl J Med 2017; 377: 1648–1656.

16.

Malmgren

Thom

. Hippocampal sclerosis—origins and imaging. Epilepsia 2012; 53: 19–33.

17.

Bast

Ramantani

Seitz

, et al. Focal cortical dysplasia: prevalence, clinical presentation and epilepsy in children and adults. Acta Neurol Scand 2006; 113: 72–81.

18.

Urbach

Kellner

Kremers

, et al. MRI of focal cortical dysplasia. Neuroradiology 2022; 64: 443–452.

19.

Kimura

Shioya

Saito

, et al. Radiologic and pathologic features of the transmantle sign in focal cortical dysplasia: the T1 signal is useful for differentiating subtypes. Am J Neuroradiol 2019; 40: 1060–1066.

20.

Kim

Park

S-H

, et al. Comparison of MRI features and surgical outcome among the subtypes of focal cortical dysplasia. Seizure 2012; 21: 789–794.

21.

Jayalakshmi

Vooturi

Vadapalli

, et al. Predictors of surgical outcome in focal cortical dysplasia and its subtypes. J Neurosurg 2022; 136: 512–522.

22.

Leventer

Jansen

Pilz

, et al. Clinical and imaging heterogeneity of polymicrogyria: a study of 328 patients. Brain 2010; 133: 1415–1427.

23.

Curry

Lammer

Nelson

, et al. Schizencephaly: heterogeneous etiologies in a population of 4 million California births. Am J Med Genet 2005; 137A: 181–189.

24.

George

Vassar

, et al. Fetal MRI findings, etiology, and outcome in prenatally diagnosed Schizencephaly. Am J Neuroradiol 2025; 46: 800–807.

25.

Nabavizadeh

Zarnow

Bilaniuk

, et al. Correlation of prenatal and postnatal MRI findings in Schizencephaly. Am J Neuroradiol 2014; 35: 1418–1424.

26.

Guerrini

Dobyns

. Malformations of cortical development: clinical features and genetic causes. Lancet Neurol 2014; 13: 710–726.

27.

Raza

Chen

Chansysouphanthong

, et al. The clinical and imaging features of gray matter heterotopia: a clinical analysis on 15 patients. Neurol Sci 2019; 40: 489–494.

28.

Oegema

Barkovich

Mancini

GMS

, et al. Subcortical heterotopic gray matter brain malformations: classification study of 107 individuals. Neurology 2019; 93: e1360–e1373.

29.

Devisme

Bouchet

Gonzalès

, et al. Cobblestone lissencephaly: neuropathological subtypes and correlations with genes of dystroglycanopathies. Brain 2012; 135: 469–482.

30.

Bican Demir

Eser

Bekar

, et al. A rare cause of epilepsy: ulegyria revisited in a series of 10 patients. Clin EEG Neurosci 2022; 53: 153–159.

31.

Stern

Elsingergy

Venkatakrishna

SSB

, et al. Frequency of ulegyria on delayed MRI scans in children with term hypoxic-ischemic injury. Pediatr Radiol 2023; 53: 104–111.

32.

Gökçe

Çevik

. Evaluation of different ulegyria patterns with magnetic resonance imaging. J Clin Neurosci 2018; 58: 148–155.

33.

Ybarra

Hafiz

Robinson

M-E

, et al. A new imaging entity consistent with partial ectopic posterior pituitary gland: report of six cases. Pediatr Radiol 2020; 50: 107–115.

34.

El Sanharawi

Tzarouchi

Cardoen

, et al. High-resolution heavily T2-weighted magnetic resonance imaging for evaluation of the pituitary stalk in children with ectopic neurohypophysis. Pediatr Radiol 2017; 47: 599–605.

35.

Côté

Salzman

Sorour

, et al. Normal dimensions of the posterior pituitary bright spot on magnetic resonance imaging: clinical article. J Neurosurg 2014; 120: 357–362.

36.

Cullingford

Abraham

Siafarikas

, et al. Septo-optic dysplasia and gastroschisis: trends in birth prevalence and association with maternal age. Eur J Pediatr 2025; 184: 297.

37.

Blackburn

Thomas

Hughes

, et al. Neuropathology of septo-optic dysplasia: a report of 4 autopsy cases. J Child Neurol 2021; 36: 105–115.

38.

Ward

Connolly

DJA

Griffiths

. Review of the MRI brain findings of septo-optic dysplasia. Clin Radiol 2021; 76: 160.e161.

39.

BWY

Kulkarni

Rutka

, et al. Clinical predictors of developmental outcome in patients with cephaloceles: clinical article. J Neurosurg: Pediatrics PED 2008; 2: 254–257.

40.

Perez Da Rosa

Millward

Bhatti

, et al. MRI findings of intracranial anomalies associated with cephalocele—a case series. Childs Nerv Syst 2014; 30: 891–895.

41.

Northrup

Aronow

Bebin

, et al. Updated international Tuberous sclerosis complex diagnostic criteria and surveillance and management recommendations. Pediatr Neurol 2021; 123: 50–66.

42.

Wang

Segaran

Bhalla

, et al. Tuberous sclerosis: current update. Radiographics 2021; 41: 1992–2010.

43.

Yeom

Comi

. Updates on Sturge-Weber Syndrome. Stroke 2022; 53: 3769–3779.

44.

Sudarsanam

Ardern-Holmes

. Sturge–Weber syndrome: from the past to the present. Eur J Paediatr Neurol 2014; 18: 257–266.

45.

Kim

Baek

, et al. Incidence of Sturge-Weber Syndrome and risk of secondary glaucoma: a nationwide population-based study using a rare disease registry. Am J Ophthalmol 2023; 247: 121–126.

46.

Baertling

Rodenburg

Schaper

, et al. A guide to diagnosis and treatment of Leigh syndrome. J Neurol Neurosurg Psychiatr 2014; 85: 257–265.

47.

Ball

Thorburn

Rahman

. Mitochondrial DNA-associated Leigh syndrome spectrum. GeneReviews®[Internet]. 2024 May 9.

48.

Bonfante

Koenig

Adejumo

, et al. The neuroimaging of Leigh syndrome: case series and review of the literature. Pediatr Radiol 2016; 46: 443–451.

49.

Bien

Tiemeier

Sassen

, et al. Rasmussen encephalitis: incidence and course under randomized therapy with tacrolimus or intravenous immunoglobulins. Epilepsia 2013; 54: 543–550.

50.

Pradeep

Sinha

Saini

, et al. Evolution of MRI changes in Rasmussen’s encephalitis. Acta Neurol Scand 2014; 130: 253–259.

51.

Thy

de Montmollin

Bouadma

, et al. Severe meningoencephalitis: epidemiology and outcomes. Curr Opin Crit Care 2023; 29: 415–422.

52.

Abbuehl

Branca

Ungureanu

, et al. Magnetic resonance imaging in acute meningoencephalitis of viral and unknown origin: frequent findings and prognostic potential. Front Neurol 2024; 15: 1359437.

53.

Dubey

Pittock

Kelly

, et al. Autoimmune encephalitis epidemiology and a comparison to infectious encephalitis. Ann Neurol 2018; 83: 166–177.

54.

Vaišvilas

Ciano-Petersen

Macarena Villagrán-García

, et al. Paraneoplastic encephalitis: clinically based approach on diagnosis and management. Postgrad Med J 2022; 99: 669–678.

55.

Kelly

Grant

Murchison

, et al. Magnetic resonance imaging characteristics of LGI1-Antibody and CASPR2-Antibody Encephalitis. JAMA Neurol 2024; 81: 525–533.

56.

Gillon

Chan

Chen

, et al. MR imaging findings in a large population of autoimmune encephalitis. Am J Neuroradiol 2023; 44: 799–806.

57.

Zhang

Henning

Zou

, et al. Intracranial ganglioglioma: clinicopathological and MRI findings in 16 patients. Clin Radiol 2008; 63: 80–91.

58.

Luzzi

Elia

Del Maestro

, et al. Dysembryoplastic Neuroepithelial Tumors: what you need to know. World Neurosurg 2019; 127: 255–265.

59.

Nunes

Hsu

da Rocha

, et al. Multinodular and vacuolating neuronal tumor of the cerebrum: a New “Leave Me Alone” lesion with a characteristic imaging pattern. Am J Neuroradiol 2017; 38: 1899–1904.

60.

Flemming

Kumar

Lanzino

, et al. Baseline and evolutionary radiologic features in sporadic, hemorrhagic brain cavernous malformations. Am J Neuroradiol 2019; 40: 967–972.

61.

Smith

. Cavernous malformations of the central nervous system. N Engl J Med 2024; 390: 1022–1028.

62.

Akers

Al-Shahi Salman

Awad I

, et al. Synopsis of guidelines for the clinical management of cerebral cavernous malformations: consensus recommendations based on systematic literature review by the angioma alliance scientific advisory board clinical experts panel. Neurosurgery 2017; 80: 7.

63.

Bonm

Ritterbusch

Throckmorton

, et al. Clinical Imaging for diagnostic challenges in the management of Gliomas: a review. J Neuroimaging 2020; 30: 139–145.