Annals of Indian Academy of Neurology
  Users Online: 5897 Home | About the Journal | InstructionsCurrent Issue | Back IssuesLogin      Print this page Email this page  Small font size Default font size Increase font size

Table of Contents
Year : 2022  |  Volume : 25  |  Issue : 6  |  Page : 1019-1028

An MRI based ischemic stroke classification – A mechanism oriented approach

Department of Neurology, Government Medical College Kozhikode, Kozhikode, Kerala, India

Date of Submission21-Apr-2022
Date of Decision11-Sep-2022
Date of Acceptance12-Sep-2022
Date of Web Publication17-Nov-2022

Correspondence Address:
Joe James
Department of Neurology, Government Medical College Kozhikode, Kozhikode, Kerala
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/aian.aian_365_22

Rights and Permissions



Oxfordshire Community Stroke Project and Trial of Org 10172 in acute stroke treatment are the commonly used ischemic stroke classification systems at present. However, they underutilize the newer imaging technologies. Diffusion-weighted magnetic resonance imaging (DW-MRI) of the brain can detect the site and extent of infarcts accurately. From the MRI patterns, the mechanisms of ischemic stroke can be inferred. We propose to classify ischemic infarcts into the following types based on their DW-MRI appearance: cortical territorial infarcts, striatocapsular infarcts, superficial perforator infarcts, cortical and deep watershed infarcts, lacunar infarcts, long insular artery (LIA) infarcts, branch atheromatous disease (BAD) infarcts, corpus callosal infarcts, infratentorial infarcts, and unclassifiable infarcts. This DW-MRI-based classification of ischemic stroke is easy, fast, and mechanism oriented. A review of the literature reveals that cortical territorial, striatocapsular, and corpus callosal infarcts are associated with embolic sources and large artery intracranial atherosclerosis. Superficial perforator and LIA infarcts are also probably embolic. Watershed infarcts are frequently associated with severe carotid disease with microembolism or hemodynamic failure. Mechanisms of BAD infarcts include microatheroma, junctional plaque or a plaque within a parent artery blocking the orifice of a large, deep penetrating, or circumferential artery. Small lacunar infarcts are due to the lipohyalinosis of penetrating arteries. Types and mechanisms of infratentorial infarcts are similar to supratentorial infarcts. Such a classification system is useful for prognosticating acute stroke, arranging specific investigations, and planning strategies for secondary prevention and research.

Keywords: Cerebrovascular accident, magnetic resonance imaging, lacunar stroke, carotid stenosis, intracranial arteriosclerosis

How to cite this article:
Jose J, James J. An MRI based ischemic stroke classification – A mechanism oriented approach. Ann Indian Acad Neurol 2022;25:1019-28

How to cite this URL:
Jose J, James J. An MRI based ischemic stroke classification – A mechanism oriented approach. Ann Indian Acad Neurol [serial online] 2022 [cited 2023 Feb 6];25:1019-28. Available from:

   Introduction Top

There are many classification systems used to subtype ischemic strokes. Several pathological and radiological studies have shown a definite relationship between ischemic stroke subtypes and stroke mechanisms.[1],[2] The clinical presentations, treatment decisions, prognosis, and the risk of recurrence of stroke also depend on the underlying mechanisms. The two commonly used stroke classifications are the Oxfordshire Community Stroke Project (OCSP) classification and Trial of ORG 10172 in Acute Stroke Treatment (TOAST) classification.

OCSP classification, introduced in 1991, categorizes stroke syndromes based on the affected vascular territory into four subtypes: total anterior circulation infarcts, partial anterior circulation infarcts, lacunar infarcts, and posterior circulation infarcts.[3] The classification was introduced for a population-based epidemiological study and was based on clinical findings. Computed tomography (CT) scan head was the investigation performed. There are many drawbacks to the OCSP classification. An approach to stroke based only on clinical and CT scan head is outdated in modern stroke diagnosis and management. The site and extent of the infarct cannot be assessed from this classification. Diffusion-weighted magnetic resonance imaging (DW-MRI) studies have shown that OCSP classification does not accurately discriminate between lacunar and small volume cortical infarcts.[4]

Since its introduction in 1993, TOAST classification has become the most widely used system for stroke classification. TOAST classification is an etiology-based classification where stroke is divided into five subtypes: large artery atherosclerosis, cardioembolism, small vessel occlusion, stroke of other determined etiology, and stroke of undetermined etiology.[5] However, TOAST classification dumps a good proportion of stroke in the group of undetermined etiology. In this classification, the diagnosis of cardioembolic stroke is based on the presence of a cardiac source of embolism, which is further classified into high and medium risk. The risk factors may be an incidental finding rather than the real cause of stroke, more so in the latter. Furthermore, the small vessel occlusions (lacunar infarcts) are defined by the clinical syndrome and size of the infarct. A small deep infarct due to other pathologies [e.g., internal water shed infarcts and superficial perforator infarcts] may be classified as a lacunar infarct. Mudden et al.[6] reported that the initial subtyping of stroke by TOAST matched the final diagnosis in only 62% of cases.

In OCSP and TOAST classifications, anatomical localization of the infarct, identification of the vascular territory, and understanding the mechanism of the stroke by clinical history, physical examination, and basic investigations are skilled works and the reliability depends on the experience of the physicians. These factors point to the need for alternative methods of stroke identification and classification based on more advanced investigations. DW-MRI can detect infarcts within minutes of stroke onset and the extent and site of infarcts can be accurately assessed. MRI and DWI are now available in almost all stroke care centers and the time taken for a DW-MRI is <3-min. If a magnetic resonance angiogram (MRA) is added to the protocol, the diagnostic accuracy still increases. In a study, it was found that pre-MRI TOAST matched the final diagnosis in 48%, which improves to 83% after DWI and 94% after DW-MRI with MRA. Pre-MRI OCSP diagnosis matched the final diagnosis in 67%, improving to 100% after DW-MRI.[7] Rovira et al.[8] has suggested a stroke classification based on DW-MRI. However, it does not define the subtypes clearly. Even though the cortical infarcts are correctly depicted, the subcortical infarcts are dumped together without clear demarcation.

In this review, we attempt to subclassify to subclassify the ischemic strokes based on the patterns of DW-MRI. The definitions are adopted from the standard definitions used in clinical studies and trials. The emphasis of the classification is to identify the stroke type by DW-MRI and to reach a logical conclusion of the stroke mechanism from the type of infarct.

We propose to classify ischemic stroke into the following types:

  1. Cortical territorial infarcts
  2. Striatocapsular infarcts (deep territorial infarcts)
  3. Superficial perforator infarcts
  4. Watershed infarcts

    1. Cortical watershed
    2. Internal watershed

  5. Long insular artery infarcts
  6. Branch atheromatous disease infarcts
  7. Lacunar infarcts
  8. Corpus callosal infarcts
  9. Infratentorial infarcts
  10. Unclassifiable infarcts.

   Search Methodology Top

We searched PubMed and google scholar for articles published upto February 2019, using the keywords 'stroke classification,' 'magnetic resonance imaging and stroke,' 'diffusion imaging and stroke,' 'territorial infarct,' 'striatocapsular infarct,' 'superficial perforator,' 'watershed infarct,' 'cortical watershed infarct,' 'lacune,' 'lacunar stroke,' 'long insular artery,' 'branch atheromatous disease,' 'subcortical infarct,' 'corpus callosal infarct,' 'pontine stroke,' midbrain stroke,' 'medulla and stroke,' 'cerebellum and stroke', 'thalamus and stroke,' 'vertebrobasilar insufficiency,' carotid stenosis,' 'intracranial arteriosclerosis,' and 'embolic stroke.' From the results, title and abstracts of the articles were screened and if found relevant, full texts were reviewed.

   Results Top

Cortical territorial infarcts

Cortical territorial infarcts are ischemic lesions involving the cerebral cortex and subcortical white matter in the territory of the major cerebral arteries, i.e., anterior cerebral arteries (ACA), middle cerebral arteries (MCA), and posterior cerebral arteries (PCA). The terminal branches of the main cerebral arteries form a pial plexus. In humans, the cerebral cortex and the subcortical U-fibers are supplied by short arterioles of less than 50-μm diameter from the brain surface, whereas the centrum semiovale is supplied by 2 to 5 cm long medullary end arteries arising from the pial plexus.[9] Occlusion of the major cerebral vessels cause ischemic lesions of the cortex and subcortical areas. These ischemic lesions can also be called pial territory infarcts.

The templates depicting the major arterial territories of the brain, introduced by Damasio or the topographic brain atlas introduced by Kim et al., can be used for the identification of territorial infarcts.[10],[11] The twelve templates of Damasio are based on diagrams of consecutive CT scan sections of the brain approximately 8 mm apart. The topographic brain atlases of Kim are digital maps of supratentorial infarcts, generated using DW-MRI. According to these templates and maps, the anterior part of the medial cortex and subcortex up to the parieto-occipital sulcus, the medial orbital gyri on the orbitofrontal surface and the superior frontal gyrus constitute the cortical territories of ACA. The lateral cortices of frontal, temporal, and parietal lobes except for a small anterior and posterior area, constitute the cortical territories of MCA. Anteriorly this extends to the superior frontal sulcus, and posteriorly to the middle occipital gyrus. The lateral orbital gyri of the inferior frontal surface are also supplied by MCA. The retrosplenial medial cortex, occipital pole and the adjacent lateral surface, and the inferomedial temporal lobe constitute the cortical territories of PCA.

Cortical territorial infarcts are seen in MRI as signal alterations confined to the areas supplied by the MCA, ACA, and PCA [[Figure 1], panels a-c]. The infarcts may be restricted to the cortex or may involve the subcortex also. Any infarct which affects the cortical ribbon and is 10 mm or more in size with or without subcortical infarcts can be taken as a cortical territorial infarct. However, cortical lesions <10 mm may represent the cortical spotty lesions of superficial perforator infarct (described later). MCA cortical territorial infarcts may be associated with lenticulostriate territory infarcts to produce a complete territorial infarct [[Figure 1], panel d]. Even though prognostically different, all these infarcts can be categorized as territorial infarcts because the mechanism of stroke in these conditions is the same.
Figure 1: DW-MRI of territorial, striatocapsular, superficial peroforator and watershed infarcts. (a) MCA Cortical territorial infarct (b) ACA Cortical territorial infarct (c) PCA territorial infarct (d) Complete MCA territorial infarct (e) Fragmented MCA territorial infarct (f) 'comma' shaped striatocapsular infarct involving the head of caudate nucleus, putamen and anterior limb of internal capsule sparing the globus pallidus, genu and posterior limb of internal capsule (g) superficially located, oval lesions of SPI scattered in the centrum semiovale associated with a spotty cortical lesion (arrow) (h) Anterior watershed infarct appearing as a wedge extending from the anterior horn of the lateral ventricle to the frontal cortex. (i) Posterior watershed infarct appearing as a wedge extending from the occipital horn of the lateral ventricle to the parieto-occipital cortex. (j) Confluent large cigar-shaped deep watershed infarct. (k) “rosary-like” deep WSI in the centrum semi ovale

Click here to view

Cortical territorial infarcts may sometimes become fragmented and appear as several small, disseminated lesions in cortical and subcortical areas [[Figure 1], panel e].[8] This may be due to the breaking up of emboli with reperfusion or due to multiple emboli. The wide variability in the territorial supply of large arteries may also create some confusion in delineating the territory of infarct in DW-MRI. However, as the mechanism of stroke in all the major cortical territorial infarctions is the same, this disparity is not important from a classification point of view.

Cortical territorial infarcts are usually caused by embolic stroke—either cardiac or artery to artery embolism from extracranial large vessels. Cardiac embolism is more common than carotid artery disease.[12],[13] However, in Asians and black population intracranial atherosclerosis is a common etiology for cortical territorial infarcts.[14],[15] Patients with cardioembolic infarcts are likely to have total anterior circulation infarcts, with a higher baseline National Institutes of Health Stroke Scale (NIHSS) score. On the other hand, patients with intracranial atherosclerosis are likely to have partial anterior circulation infarcts with lower baseline NIHSS score and milder neurological deficits.[16]

Striatocapsular (Deep territorial) infarcts

Striatocapsular infarcts (SCI) are a distinct form of subcortical infarcts in the striatocapsular area caused by simultaneous occlusion of more than one adjacent lenticulostriate arteries. Lenticulostriate arteries are deep perforators that arise from the proximal part of MCA. These deep perforating branches supply the head and body of caudate nucleus (superior part), lateral segment of globus pallidus, the putamen, dorsal half of internal capsule and the lateral part of the anterior commissure. The size, shape, site, pathogenesis and clinical features distinguish SCI from other subcortical infarcts.

The first systematic description of SCI was given by Bladin et al. in 1984.[1] They found 11 cases of SCI in 1600 patients admitted for stroke in the Austin Hospital Stroke Unit. The definition of SCI is based on its radiological characteristics.[17] On axial CT/MRI, these infarcts are lentiform, triangular or 'comma'-shaped [[Figure 1], panel f]. The size of the infarct is 3 – 4.5 cm, with a width of 1 – 2 cm and a depth of 2 – 4 cm. The infarct involves the head of caudate nucleus, putamen and anterior limb of internal capsule. Globus pallidus, genu and posterior limb of internal capsule are usually spared. The overlying cortex is also spared. The typical comma-shaped lesion has a head as caudate nucleus along with anterior limb of internal capsule and a tail as lentiform nucleus.{Figure 1}

Some studies have included coma, lenticular or triangular lesion which involves at least two components in the striatocapsular area—head of caudate plus internal capsule or putamen plus internal capsule also as SCI.[18] There are some controversies regarding the size of the SCI also. Even though 3 cm is the most common accepted size, some studies used a lower limit of 2 cm.[18]

The most common clinical manifestation of SCI is weakness.[17],[19] Hemiplegia is due to the involvement of the corticospinal tract in the posterosuperior segment of the lenticulostriate artery (LSA) territory.[20] LSA territory can be divided into a superior and an inferior subsegment in the coronal plane. This can be further divided into an anterior and a posterior subsegment in the axial plane. The anterior segment includes the superior part of the head of caudate nucleus, anterior limb of the internal capsule and anterior putamen. The posterior segment includes the body of caudate nucleus, posterior putamen and superior part of posterior limb of internal capsule. Konishi et al.[20] by MR tractography have demonstrated that the corticospinal tract crosses the LSA territory only at the posterosuperior subsegment and corticospinal tract involvement in this subsegment has a significant correlation with stroke severity. As the corticospinal tract descends, it quickly exits the LSA territory and enters the area supplied by the anterior choroidal artery. This may be the reason for the radiological sparing of the posterior limb of internal capsule in many SCI, even though hemiplegia is the commonest clinical manifestation. Another important feature of striatocapsular infarct is the presence of cortical signs such as dysphasia, dyspraxia, hemineglect, and eye deviation.[19]

Identification of SCI is important because the mechanism of stroke in SCI is different from lacunar infarcts. Cardiac abnormalities and severe carotid artery disease sufficient enough to produce an embolism to the M1 segment of MCA and intracranial atherosclerotic disease affecting the MCA predisposing to LSA ostial thrombosis are the important cause of SCI.[19],[21]

Superficial perforator infarcts

Acute infarctions confined to the territory of the superficial perforator arteries [white matter medullary arteries] are called superficial perforator infarcts (SPI). Superficial perforator arteries originate from the pial branches of anterior, middle, and posterior cerebral arteries. They are 2- – 5-cm long end arteries that descend toward the upper part of lateral ventricle and supply the white matter of centrum semiovale.[9]

An SPI (white-matter medullary infarct) can be defined as an infarct located in the territory of the perforating medullary artery. They are superficially located, oval or circular lesions scattered in the centrum semiovale and may be associated with spotty cortical lesions [[Figure 1], panel g]. Radiologically, the outermost limit of SPI is taken as the cortical ribbon, while the innermost limit is taken as the corona radiata at the level of the deep perforating artery.[22]

Spotty cortical lesions are defined as small hyperintense signals in the cortex <10 mm in size detected by DW-MRI that are smaller than lesions of white matter medullary infarcts.[23],[24] Spotty cortical lesions detected by DW-MRI are associated with microembolic signals in transcranial doppler studies and are related to small infarcts due to microemboli originating from the heart or large arteries occluding the small cortical arteriols.[24]

In DW-MRI, it may be difficult to differentiate between internal watershed infarcts and SPI. Hence, some studies have lumped them together and categorized them as subcortical white matter infarcts. Lee et al.[22] in a study of 54 patients with SPI and 29 patients with internal watershed infarcts, found that SPI were superficially located, oval or circular lesions and were widely scattered, whereas internal watershed infarcts showed a tendency to localize on paraventricular regions where it appeared as a chain-like or sausage-like lesions. The diameter of the internal watershed infarct was significantly larger than SPI. They also found that SPI are frequently associated with cortical spotty lesions. SPI usually has a lower NIHSS score and a favorable outcome.

The pathogenesis and etiology of SPI are not definite. However, the most common pathology considered is embolic. Lammie et al.[25] in a postmortem study of 12 cases of small centrum semiovale infarcts found that 10 out of 12 cases had probable embolic etiology. Yonemura et al.[26] also found that small centrum ovale lesions were associated with large-vessel and heart diseases. Boiten et al.[27] in a sub-analysis of the European Carotid Surgery Trial found that small white matter medullary infarcts were associated with carotid large artery disease in 66% of cases. Lee et al.[23] in a study of 103 patients with medullary infarcts found that 65 patients (63%) had large artery disease and 12 (11.7%) had cardiac embolic sources.[23] More than 80% of them had cortical spotty lesions indicating an embolic mechanism.

Watershed infarcts

Watersheds are areas that lie at the junction of two different drainage areas. Watershed infarcts (WSI) are ischemic lesions that occur in characteristic locations at the junction of two non-anastomosing arterial territories.

There are two supratentorial watershed areas.

  1. Cortical watershed area (external watershed): These are the areas between the cortical territories of ACA, MCA, and PCA. The watershed area between ACA and MCA is called the anterior cortical watershed area and that between MCA and PCA is called the posterior cortical watershed area.
  2. Deep watershed area (internal watershed): These are areas between the territories of deep and superficial (medullary) perforating arteries. Superficial perforating arteries are 2 – 5 cm long end arteries arising from the pial plexus of middle cerebral, anterior cerebral, and posterior cerebral arteries. Deep watershed areas are formed between these medullary arteries and the Heubner's, lenticulostriate, and anterior choroidal arteries. Anatomically this area lies in the white matter along and slightly above the lateral ventricle.

WSI are best demonstrated by DW-MRI in the acute phase. An infarct is considered to be in a watershed area when the border between two main arterial territories divides the infarct into two parts so that the smallest part would be at least one-third of the total infarct.[28] This criterion has to be fulfilled in all radiological slices in which infarcts are seen. The territories of cerebrovascular supply can be demarcated using the templates of the topographic brain atlas of Kim.[11]

Radiologically, anterior WSI appear as a fronto-parasagittal wedge extending from the anterior horn of the lateral ventricle to the frontal cortex [[Figure 1], panel h] or as a linear strip in the paramedian white matter slightly lateral to the interhemispheric fissure. Posterior WSI appear as parieto-temporo-occipital wedge extending from the occipital horn of the lateral ventricle to the parieto-occipital cortex [[Figure 1], panel I].[29]

Radiologically, internal WSI can be of two types – confluent and partial.[30] Confluent infarcts are larger and cigar-shaped [[Figure 1], panel J]. Partial infarcts have a 'Rosary-like' (chain-like) appearance ([Figure 1], panel K). Both are seen alongside the lateral ventricle and in the centrum semi ovale.[31] Rosary-like internal WSI may be defined as three or more lesions 3 mm or greater in diameter arranged in a linear pattern parallel to the lateral ventricle in the centrum semiovale or corona ratiata.[32]

There are several studies focusing on the pathophysiology and mechanisms of WSI. These infarcts are associated with large artery atherosclerosis and cardioembolic sources. Microembolism and hemodynamic failure due to hypotension are the main mechanisms postulated. Microemboli are small emboli of 50–300 μm in size, mostly composed of cholesterol crystals.[33] They arise from unstable carotid plaques or the stump of an occluded internal carotid artery. Small thrombi travel preferentially to watershed areas because of their smaller size. Internal WSI, especially partial type, are more associated with hemodynamic mechanisms in patients with severe carotid disease, whereas cortical WSI are more associated with cardiac disease and embolism.[29],[31],[32],[34]

Long insular artery infarcts

The long insular artery (LIA), which arises from the insular segment of MCA (M2), has been anatomically recognized as a subtype of the white matter medullary artery. It supplies the insular cortex, extreme capsule, claustrum, and external capsule, often extending to the corona radiata.[35]

LIA infarcts are frequently mistaken as lenticulostriate infarcts (lacunar infarcts) because the sizes and shapes of both infarcts are similar in axial MR imaging. The LIA infarcts are best identified in coronal MRI. The subcortical white matter and basal ganglia on coronal MRI images can be divided into three vascular territories: The white matter medullary arteries (WMMA) territory, the long insular arteries (LIA) territory, and the lenticulostriate arteries territory [Figure 2]. A virtual line from the tip of the anterior horn of lateral ventricle to the top of the superior limb of the insular cleft corresponds closely to the vascular territory of the LIA and an infarct along this line can be considered a LIA infarct [Figure 2].[36] Those infarcts situated under this line and extending vertically (craniocaudal) can be considered lenticulostriate artery infarcts. Patients with LIA infarctions demonstrate classic lacunar syndromes. The prevalence of embolic high-risk sources and moderate-risk sources is significantly higher in the long insular artery than in the lenticulostriate artery group.[2]
Figure 2: Vascular territories of perforating arteries and their infarct patterns. (A) vascular territories of the white matter medullary arteries (WMMA), the long insular arteries (LIA) and the lenticulostriate arteries (LSA) M1-horizontal segment of MCA, M2- insular segment of MCA, M3-opercular segment of MCA, M4-cortical segment of MCA. (B) Infarct site and shape in perforator infarcts in coronal plain. (a) white matter medullary artery infarct. (b) LIA infarct seen as an infarct extending from the tip of the anterior horn of the lateral ventricle to the top of the superior limb of the insular cleft. (c) LSA infarct seen as a vertically extending (craniocaudal) infarct in the gangliocapsular area

Click here to view

Branch atheromatous disease

Branch Atheromatous Disease (BAD) is a pathological finding of stenosis or occlusion at the origin of a large size penetrating artery, due to a microatheroma or a large parent artery plaque. The term BAD was introduced by Caplan to explain an alternative mechanism other than lipohyalinosis for small subcortical infarcts.[1] BAD affects arteries of larger caliber like large proximal lenticulostriate arteries, basilar artery branches, Heubner's artery, anterior choroidal arteries and thalamogeniculate arteries.[37],[38],[39] There are three mechanisms of stroke in BAD - plaque within a parent artery blocking the branch orifice, plaque extending into the branch from the parent artery (junctional plaque) and microatheroma originating in the orifice of a branch [[Figure 3], panels a-c].[37] Current imaging modalities such as MR angiography, CT angiography, and DSA are of limited utility in the diagnosis of BAD. They depict the morphology of arteries and are unable to show inner vessel wall changes. High-resolution 3-Tesla magnetic resonance imaging has recently been used for visualization of the inner wall of MCA and basilar artery and is useful in the diagnosis of BAD.[38],[39]
Figure 3: Branch atheromatous disease infarct. Mechanisms of BAD infarct include (a) plaque within a parent artery blocking the branch orifice, (b) plaque extending into the branch from the parent artery (junctional plaque) (c) microatheroma originating in the orifice of a branch. DW-MRI of BAD infarct in lenticulostriate artery showing an infarct in the gangliocapsular area (d), extending down in consecutive cuts upto the origin of LSA (e and f). (g) Branch atheromatous disease infarct

Click here to view

A recent review searching the available literature showed a lack of clear-cut definitions of BAD strokes.[40] Even though, the term BAD implies an arterial pathology, most studies of BAD are based on vascular territory, size, and shape of the infarcts. Yamamoto et al.[41] defined BAD in lenticulostriate artery territory as infarcts with a size more than 10 mm in diameter and visible in three or more axial slices at 7 mm and that of anterior pontine arteries as unilateral infarcts extending to the basal surface of the pons on MRI. The STRIVE Classification of “small subcortical infarct” with a size less than 20 mm does not differentiate between BAD and lacunar infarcts.[42] Nakase et al.[43] defined BAD infarct as a subcortical lesion ≥15 mm in diameter in more than 3 slices at 5 mm [[Figure 3], Panels D-G] or a lesion extending to the surface of the pontine base on DW-MRI.

As the BAD vascular lesions are located proximally along the perforator arteries, BAD infarcts are larger and have a poorer prognosis than lacunar infarcts. Compared to lacunar strokes, the duration of hospitalization and the residual disability of patients are also significantly greater, as is early neurological deterioration. Early neurological deterioration, defined as an increase of more than 2-point in the National Institutes of Health Stroke Scale within 48 h of stroke onset, is a well-known phenomenon in BAD stroke.[43]

It is found that BAD is associated with intracranial atherosclerotic disease of MCA and basilar artery.[44],[45]

Lacunar Infarcts

Lacunar infarcts are small infarcts in the basal ganglia, thalamus, brainstem (especially pons), internal capsule, and deep cerebral white matter resulting from the occlusion of a single small perforating artery. Acute lacunar infarcts are seen in DW-MRI as small hyperintense signal lesions of <15-mm size in the classical sites of these infarcts [[Figure 4], panel a]. The terms lacune, lacunar stroke and lacunar infarct are not the same. A lacune is a small fluid-filled cavity that is considered the healed stage of a small deep brain infarct. Lacunar stroke is a clinical stroke syndrome with the symptoms and signs of a small subcortical or brainstem lesion. Lacunar infarct is a clinical stroke syndrome of the lacunar type where the underlying lesion on brain imaging is an infarct.
Figure 4: DW-MRI of lacunar infarct and infratentorial infarcts. (a) Lacunar infarct seen in DW-MRI as a small hyperintense signal lesion of < 15-mm size in the gangliocapsular area (b) BAD infarct of anterior pontine artery seen as unilateral paramedian infarct of the pons extending to the basal surface. (c) Bilateral isolated infarctions of the pons showing the “heart appearance”

Click here to view

The concept of lacunar stroke, the characteristics of the infarcts, the clinical syndromes, and the arterial pathology causing lacunar infarcts were described best by Miller Fisher.[46] These infarcts are due to the occlusion of a single vessel of the perforator arterial system, i.e., lenticulostriate, thalamoperforator and paramedian arterioles of the brainstem. The most common pathology of the blood vessels producing lacunar infarcts is lipohyalinosis.[46] Lipohyalinosis is destructive segmental microangiopathy of small vessels, histologically characterized by loss of arterial architecture, vessel wall thickening, focal arteriolar dilatation, and extravasation of blood components through the wall. In acute cases, evidence of fibrinoid vessel wall necrosis is also seen. Such vascular lesions involve small arteries 40-200 μm in diameter.[46]

The TOAST classification and National Institute of Neurological Disorders and Stroke define lacunar infarcts as brain infarctions <15 mm in diameter and accompanied by a lacunar syndrome.[47] The classical lacunar syndromes are pure motor stroke, pure sensory stroke, ataxic hemiparesis, dysarthria-clumsy hand syndrome, and mixed sensorimotor syndrome. The Standards for Reporting Vascular changes on Neuroimaging (STRIVE) Criteria define lacune as a round or ovoid, fluid-filled cavity between 3 mm and 15 mm in diameter in the territory of one perforating arteriole.[42] On FLAIR-MRI images, lacunes have a central CSF-like hypointensity with a surrounding rim of hyperintensity. A typical lacune evolves from a “recent small subcortical infarct” in the territory of a deep perforating arteriole.[42]

Even though lipohyalinosis is the proposed mechanism of lacunar infarcts, other mechanisms may also be contributing to the subcortical infarcts of <15 mm in size on DW-MRI.[48],[49] In pathological studies, Fisher described not only lipohyalinosis but also microatheroma and embolism as the mechanisms of lacunar infarcts. It was found that lipohyalinosis affects vessels 40–200 μm in diameter and produces lacune of 2 – 5 mm in size. Microatheroma and emboli affect vessels 200–850 μm in diameter and produce lacune of >5 mm in size. Subsequent studies using CT/MRI have suggested the possibility of two subtypes of lacunar infarcts; one associated with white matter hyperintensities (WMH), previous lacunes, and hypertension and the other not associated with these features. It is considered that lacunar infarcts associated with WMH and asymptomatic lacunes are lipohyalinotic in origin, whereas isolated lacunar infarcts without these changes may be due to microatheroma and embolism.[48],[49]

Lacunae have to be differentiated from enlarged perivascular space which has a signal intensity of CSF. The latter is smaller than 3 mm in size and appears ovoid or round when imaged perpendicular to the course of a vessel, or linear when imaged parallel to the vessel.[42] Enlarged perivascular spaces are common in anterior perforated substances and the lower part of the basal ganglia and putamen.

Key diagnostic characteristics and etiologies of various types of infarcts are summarized in [Table 1].
Table 1: Etiologies and MRI characteristics of various types of infarcts

Click here to view

Corpus Callosal Infarcts

Corpus callosal infarcts are rare. This is because of the peculiarities of vascular supply of corpus callosum. The corpus callosum derives its blood supply from three main arterial systems, anterior communicating artery (subcallosal artery and median callosal artery) supplying the genu and rostrum of corpus callosum, pericallosal artery supplying the body, and the splenial artery from PCA supplying the splenium. These arteries form a callosal pial plexus from which short arterioles of less than 100 um diameters and 8 mm length, penetrate the white matter.[9]

Because of the peculiarities of blood supply, vascular changes with aging, hypertension, Binswanger disease and lacunar infarcts are rare in corpus callosum. However, cerebral vasculitis is one of the rare but important causes of corpus callosal infarction.

The commonest part of corpus callosum involved in infarction is splenium followed by body and genu. Rostral infarcts are rare. The main etiology for corpus callosal stroke are cardioembolism and large artery atherosclerosis.[50],[51] Thus apart from the vasculitic infarcts, the mechanism of corpus callosal infarcts and cortical territorial infarcts are similar.

Infratentorial Infarcts

The brainstem is supplied by long circumferential, short circumferential, and small perforating arteries arising from vertebral, basilar, and posterior cerebral arteries. The long circumferential arteries which include the posterior inferior cerebellar arteries (PICA), the anterior inferior cerebellar arteries (AICA), and the superior cerebellar arteries, supply the cerebellum also. Most posterior circulation strokes are characterized by the concomitant involvement of brainstem, cerebellum, thalamus, and occipital lobe and the etiology is usually embolic.[52] However, isolated small brainstem infarcts are mostly caused by BAD or lacunar infarcts. BAD infarcts are more common than lacunar infarcts. The BAD infarcts and lacunar infarcts of anterior circulation are defined by their size, whereas in the brainstem, it is identified by their site, size, and shape.

Large artery atherosclerosis is the most common cause of isolated medullary infarcts and they produce infarcts in the lateral and posterior medulla.[53] Lacunar infarcts are rare in the medulla because most of the medulla is supplied by circumferential arteries rather than by perforating arteries from vertebra arteries. They usually involve the medial and anterior parts of the medulla.[53] Isolated pontine infarct constitutes ~15% of posterior circulation strokes. Wedge-shaped paramedian infarcts that extend to the surface of the pons are attributed to basilar branch atherosclerosis [[Figure 4], panel b], whereas small well-circumscribed 'deep infarcts' are thought to be due to lipohyalinotic small vessel disease.[54] Isolated midbrain infarcts are rare. Small deep infarcts are caused by occlusion of penetrating branches from the basilar artery and are possibly considered as lacunar pathology. Infarcts extending to midbrain surface are caused by atherosclerosis of PCA.[55]

Bilateral isolated infarctions of the medial medulla and pons are rare. The characteristic brain MRI finding of these infarcts has been described as “heart appearance” on diffusion-weighted imaging [[Figure 4], panel c]. This appearance is due to bilateral involvement of the anteromedial and the anterolateral arterial territories sparing the lateral territories. The anteromedial and the anterolateral medulla and pons are supplied by paramedian and the short circumferential arteries. the lateral medulla and pons are supplied by the long circumferential branches, namely posterior inferior cerebellar artery, anterior inferior cerebellar artery, and superior cerebellar artery. Large-artery atherosclerosis and branch disease are the most common stroke mechanisms in such infarcts.[56],[57]

Isolated cerebellar infarcts can be territorial or non-territorial. Territorial infarcts are large infarcts in the territory of long circumferential arteries. The mechanisms of these infarcts are embolic or atherosclerotic in situ thrombosis of basilar or long circumferential arteries.[58],[59] Cerebellar infarcts <2 cm are considered borderzone infarcts (non-territorial).[60] However, the mechanism of stroke in non-territorial and territorial cerebellar infarcts are same.

Unclassifiable infarcts

It may not be possible to include all the cerebral infarcts in the above categories, the reasons being delay in getting a DW-MRI, forme fruste infarcts, ambiguity in the definitions of some subcortical infarcts and inexperience of the physicians. These infarcts can be included under unclassifiable infarcts.

This classification system has certain limitations. Since this is an MRI-bases classification, its utility may be limited in small centers and community-based stroke programs where MRI is unavailable. Reperfusion therapy may alter the MRI pattern of an infarct. For example, a large striatocapsular infarct may become smaller in size after thrombolysis and may mimic a BAD infarct or a lacunar infarct. This may result in over-or underrespresentation of stroke types and may result in misunderstanding of stroke mechanisms. Difficulty may be encountered in classifying subcortical infarcts correctly based on their size, due to lack of uniform definitions. Lastly, further studies are needed to identify the MRI patterns of 'infarcts of other etiologies' and 'infarcts of unknown etiology' described in the TOAST classification.

   Conclusions Top

To conclude, DW-MRI, which is available in most stroke centres, can be used for classifying strokes more scientifically and objectively. DW-MRI patterns can predict the possible stroke etiologies and mechanisms accurately. Further target-oriented investigations, such as angiograms, high-resolution magnetic resonance imaging, and cardiac evaluation can bring out the stroke mechanisms more accurately. This will make the stroke workup easier and more precise. Moreover, classifying clinical phenotypes according to these patterns can influence the treatment protocols and predict the prognosis better. Above all, this classification system opens a new horizon for clinical trials related to acute ischemic stroke.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

   References Top

Jörgensen L, Torvik A. Ischaemic cerebrovascular diseases in an autopsy series Part 2. Prevalence, location, pathogenesis, and clinical course of cerebral infarcts. J Neurol Sci 1969;9:285-320.  Back to cited text no. 1
Ringelstein EB, Schneider R, Koschorke S. Analysis of patterns of hemispheric brain infarctions on CT: Embolic stroke mechanism, territorial infarctions and lacunae. Psychiatry Res 1989;29:273-6.  Back to cited text no. 2
Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. Classification and natural history of clinically identifiable subtypes of cerebral infarction. Lancet 1991;337:1521-6.  Back to cited text no. 3
Asdaghi N, Jeerakathil T, Hameed B, Saini M, McCombe JA, Shuaib A, et al. Oxfordshire community stroke project classification poorly differentiates small cortical and subcortical infarcts. Stroke 2011;42:2143-8.  Back to cited text no. 4
Adams HP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in acute stroke treatment. Stroke 1993;24:35-41.  Back to cited text no. 5
Madden KP, Karanjia PN, Adams HP, Clarke WR. Accuracy of initial stroke subtype diagnosis in the TOAST study. Trial of ORG 10172 in acute stroke treatment. Neurology 1995;45:1975-9.  Back to cited text no. 6
Lee LJ, Kidwell CS, Alger J, Starkman S, Saver JL. Impact on stroke subtype diagnosis of early diffusion-weighted magnetic resonance imaging and magnetic resonance angiography. Stroke 2000;31:1081-9.  Back to cited text no. 7
Rovira A, Grivé E, Rovira A, Alvarez-Sabin J. Distribution territories and causative mechanisms of ischemic stroke. Eur Radiol 2005;15:416-26.  Back to cited text no. 8
Moody DM, Bell MA, Challa VR. Features of the cerebral vascular pattern that predict vulnerability to perfusion or oxygenation deficiency: An anatomic study. AJNR Am J Neuroradiol 1990;11:431-9.  Back to cited text no. 9
Damasio H. A computed tomographic guide to the identification of cerebral vascular territories. Arch Neurol 1983;40:138-42.  Back to cited text no. 10
Kim D-E, Park J-H, Schellingerhout D, Ryu W-S, Lee S-K, Jang MU, et al. Mapping the supratentorial cerebral arterial territories using 1160 large artery infarcts. JAMA Neurol 2019;76:72.  Back to cited text no. 11
Hannon N, Sheehan O, Kelly L, Marnane M, Merwick A, Moore A, et al. Stroke associated with atrial fibrillation--incidence and early outcomes in the North Dublin population stroke study. Cerebrovasc Dis 2010;29:43-9.  Back to cited text no. 12
Flaherty ML, Kissela B, Khoury JC, Alwell K, Moomaw CJ, Woo D, et al. Carotid artery stenosis as a cause of stroke. Neuroepidemiology 2013;40:36-41.  Back to cited text no. 13
Inzitari D, Hachinski VC, Taylor DW, Barnett HJ. Racial differences in the anterior circulation in cerebrovascular disease. How much can be explained by risk factors? Arch Neurol 1990;47:1080-4.  Back to cited text no. 14
Sacco RL, Kargman DE, Gu Q, Zamanillo MC. Race-ethnicity and determinants of intracranial atherosclerotic cerebral infarction. The Northern Manhattan stroke study. Stroke 1995;26:14-20.  Back to cited text no. 15
Lee PH. Isolated middle cerebral artery disease: Clinical and neuroradiological features depending on the pathogenesis. J Neurol Neurosurg Psychiatry 2004;75:727-32.  Back to cited text no. 16
Bladin PF, Berkovic SF. Striatocapsular infarction: Large infarcts in the lenticulostriate arterial territory. Neurology 1984;34:1423-30.  Back to cited text no. 17
Jung S, Hwang S-H, Lee B-C. Distinct clinical expressions of striatocapsular infarction according to cortical manifestations. Eur J Neurol 2004;11:627-33.  Back to cited text no. 18
Donnan GA, Bladin PF, Berkovic SF, Longley WA, Saling MM. The stroke syndrome of striatocapsular infarction. Brain 1991;114:51-70.  Back to cited text no. 19
Konishi J, Yamada K, Kizu O, Ito H, Sugimura K, Yoshikawa K, et al. MR tractography for the evaluation of functional recovery from lenticulostriate infarcts. Neurology 2005;64:108-13.  Back to cited text no. 20
Weiller C, Ringelstein EB, Reiche W, Thron A, Buell U. The large striatocapsular infarct. A clinical and pathophysiological entity. Arch Neurol 1990;47:1085-91.  Back to cited text no. 21
Lee PH, Bang OY, Oh SH, Joo IS, Huh K. Subcortical white matter infarcts: Comparison of superficial perforating artery and internal border-zone infarcts using diffusion-weighted magnetic resonance imaging. Stroke 2003;34:2630-5.  Back to cited text no. 22
Lee PH. Pathogenesis of deep white matter medullary infarcts: A diffusion weighted magnetic resonance imaging study. J Neurol Neurosurg Psychiatry 2005;76:1659-63.  Back to cited text no. 23
Kimura K, Minematsu K, Koga M, Arakawa R, Yasaka M, Yamagami H, et al. Microembolic signals and diffusion-weighted MR imaging abnormalities in acute ischemic stroke. AJNR Am J Neuroradiol 22:1037-42.  Back to cited text no. 24
Lammie GA, Wardlaw JM. Small centrum ovale infarcts--a pathological study. Cerebrovasc Dis 9:82-90.  Back to cited text no. 25
Yonemura K, Kimura K, Minematsu K, Uchino M, Yamaguchi T. Small centrum ovale infarcts on diffusion-weighted magnetic resonance imaging. Stroke 2002;33:1541-4.  Back to cited text no. 26
Boiten J, Rothwell PM, Slattery J, Warlow CP. Frequency and degree of carotid stenosis in small centrum ovale infarcts as compared to lacunar infarcts. Cerebrovasc Dis 1997;7:138-43.  Back to cited text no. 27
Bogousslavsky J, Regli F. Borderzone infarctions distal to internal carotid artery occlusion: Prognostic implications. Ann Neurol 1986;20:346-50.  Back to cited text no. 28
Yong SW, Bang OY, Lee PH, Li WY. Internal and cortical border-zone infarction: Clinical and diffusion-weighted imaging features. Stroke 2006;37:841-6.  Back to cited text no. 29
Bladin CF, Chambers BR. Clinical features, pathogenesis, and computed tomographic characteristics of internal watershed infarction. Stroke 1993;24:1925-32.  Back to cited text no. 30
Momjian-Mayor I, Baron J-C. The pathophysiology of watershed infarction in internal carotid artery disease. Stroke 2005;36:567-77.  Back to cited text no. 31
Derdeyn CP, Khosla A, Videen TO, et al. Severe hemodynamic impairment and border zone--region infarction. Radiology 2001;220:195-201.  Back to cited text no. 32
Masuda J, Yutani C, Ogata J, Kuriyama Y, Yamaguchi T. Atheromatous embolism in the brain: A clinicopathologic analysis of 15 autopsy cases. Neurology 1994;44:1231-7.  Back to cited text no. 33
Sorgun MH, Rzayev S, Yilmaz V, Isıkay CT. Etiologic subtypes of watershed infarcts. J Stroke Cerebrovasc Dis 2015;24:2478-83.  Back to cited text no. 34
Türe U, Yaşargil MG, Al-Mefty O, Yaşargil DC. Arteries of the insula. J Neurosurg 2000;92:676-87.  Back to cited text no. 35
Tamura A, Kasai T, Akazawa K, Nagakane Y, Yoshida T, Fujiwara Y, et al. Long insular artery infarction: Characteristics of a previously unrecognized entity. Am J Neuroradiol 2014;35:466-71.  Back to cited text no. 36
Caplan LR. Intracranial branch atheromatous disease: A neglected, understudied, and underused concept. Neurology 1989;39:1246-50.  Back to cited text no. 37
Chung J-W, Kim BJ, Sohn CH, Yoon B-W, Lee S-H. Branch atheromatous plaque: A major cause of lacunar infarction (high-resolution MRI study). Cerebrovasc Dis Extra 2012;2:36-44.  Back to cited text no. 38
Klein IF, Lavallée PC, Mazighi M, Schouman-Claeys E, Labreuche J, Amarenco P. Basilar artery atherosclerotic plaques in paramedian and lacunar pontine infarctions: A high-resolution MRI study. Stroke 2010;41:1405-9.  Back to cited text no. 39
Petrone L, Nannoni S, Del Bene A, Palumbo V, Inzitari D. Branch atheromatous disease: A clinically meaningful, yet unproven concept. Cerebrovasc Dis 2016;41:87-95.  Back to cited text no. 40
Yamamoto Y, Ohara T, Hamanaka M, Hosomi A, Tamura A, Akiguchi I. Characteristics of intracranial branch atheromatous disease and its association with progressive motor deficits. J Neurol Sci 2011;304:78-82.  Back to cited text no. 41
Wardlaw JM, Smith EE, Biessels GJ, Cordonnier C, Fazekas F, Frayne R, et al. Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol 2013;12:822-38.  Back to cited text no. 42
Nakase T, Yoshioka S, Sasaki M, Suzuki A. Clinical evaluation of lacunar infarction and branch atheromatous disease. J Stroke Cerebrovasc Dis 2013;22:406-12.  Back to cited text no. 43
Lee DK, Kim JS, Kwon SU, Yoo S-H, Kang D-W. Lesion patterns and stroke mechanism in atherosclerotic middle cerebral artery disease: Early diffusion-weighted imaging study. Stroke 2005;36:2583-8.  Back to cited text no. 44
Nah H-W, Kang D-W, Kwon SU, Kim JS. Diversity of single small subcortical infarctions according to infarct location and parent artery disease: Analysis of indicators for small vessel disease and atherosclerosis. Stroke 2010;41:2822-7.  Back to cited text no. 45
Miller Fisher C. Lacunar infarcts – A review. Cerebrovasc Dis 1991;1:311-20.  Back to cited text no. 46
Special report from the National Institute of Neurological Disorders and Stroke. Classification of cerebrovascular diseases III. Stroke 1990;21:637-76.  Back to cited text no. 47
Gouw AA, van der Flier WM, Pantoni L, Inzitari D, Erkinjuntti T, Wahlund LO, et al. On the etiology of incident brain lacunes: Longitudinal observations from the LADIS study. Stroke 2008;39:3083-5.  Back to cited text no. 48
Boiten J, Lodder J, Kessels F. Two clinically distinct lacunar infarct entities? A hypothesis. Stroke 1993;24:652-6.  Back to cited text no. 49
Chrysikopoulos H, Andreou J, Roussakis A, Pappas J. Infarction of the corpus callosum: Computed tomography and magnetic resonance imaging. Eur J Radiol 1997;25:2-8.  Back to cited text no. 50
Bhole R, Rochestie D, Brown C, Arora R, Giraldo E. Acute ischemic stroke in the corpus callosum: Characteristics, risk factors, and outcomes (P03.019). Neurology 2012;78:P03.019.  Back to cited text no. 51
Caplan LR. “Top of the basilar” syndrome. Neurology 1980;30:72-9.  Back to cited text no. 52
Kim K, Lee HS, Jung YH, Kim YD, Nam HS, Nam CM, et al. Mechanism of medullary infarction based on arterial territory involvement. J Clin Neurol 2012;8:116.  Back to cited text no. 53
Erro ME, Gallego J, Herrera M, Bermejo B. Isolated pontine infarcts: Etiopathogenic mechanisms. Eur J Neurol 2005;12:984-8.  Back to cited text no. 54
Kim JS, Kim J. Pure midbrain infarction: Clinical, radiologic, and pathophysiologic findings. Neurology 2005;64:1227-32.  Back to cited text no. 55
Pongmoragot J, Parthasarathy S, Selchen D, Saposnik G. Bilateral medial medullary infarction: A systematic review. J Stroke Cerebrovasc Dis 2013;22:775-80.  Back to cited text no. 56
Kumral E, Bayülkem G, Evyapan D. Clinical spectrum of pontine infarction. J Neurol 2002;249:1659-70.  Back to cited text no. 57
Kase CS, Norrving B, Levine SR, Babikian VL, Chodosh EH, Wolf PA, et al. Cerebellar infarction. Clinical and anatomic observations in 66 cases. Stroke 1993;24:76-83.  Back to cited text no. 58
Amarenco P, Hauw JJ, Gautier JC. Arterial pathology in cerebellar infarction. Stroke 1990;21:1299-305.  Back to cited text no. 59
Amarenco P, Kase CS, Rosengart A, Pessin MS, Bousser MG, Caplan LR. Very small (border zone) cerebellar infarcts. Distribution, causes, mechanisms and clinical features. Brain 1993;116:161-86.  Back to cited text no. 60


  [Figure 1], [Figure 2], [Figure 3], [Figure 4]

  [Table 1]


Print this article  Email this article


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Article in PDF (1,067 KB)
    Citation Manager
    Access Statistics
    Reader Comments
    Email Alert *
    Add to My List *
* Registration required (free)  

     Search Methodology
    Article Figures
    Article Tables

 Article Access Statistics
    PDF Downloaded164    
    Comments [Add]    

Recommend this journal