Transcranial Doppler Ultrasound Criteria for Hemodynamically Significant Internal Carotid Artery Stenosis Based on Residual Lumen Diameter Calculated From En Bloc Endarterectomy Specimens
Background and Purpose Transcranial Doppler (TCD) is often used in conjunction with carotid duplex ultrasonography (CDUS) to evaluate the hemodynamic significance of internal carotid artery (ICA) stenosis. We examined the sensitivity and specificity of TCD criteria for detection of a hemodynamically significant stenosis (residual lumen diameter <1.5 mm) at the origin of the ICA.
Methods We selected patients who underwent carotid endarterectomy (CEA) and had preoperative TCD data available. Eighty-one patients underwent transorbital evaluation, 49 of whom also had transtemporal TCD performed. The endarterectomy specimens were removed en bloc and sectioned, and the minimal residual lumen diameter calculated by computer analysis.
Results For the transorbital approach, the strongest indicators of a residual lumen diameter <1.5 mm were reversed flow in the ipsilateral ophthalmic artery and a >50% peak systolic velocity difference between the carotid siphons (distal ICAs) in patients with unilateral ICA origin stenosis. They were 100% specific and 31% and 26% sensitive, respectively. For the transtemporal approach in patients with a unilateral stenosis, a >35% difference in ipsilateral middle cerebral artery (MCA) peak systolic velocity relative to the contralateral MCA or a >50% difference in contralateral anterior cerebral artery (ACA) peak systolic velocity relative to the ipsilateral ACA were 100% specific for identifying a residual lumen diameter of <1.5 mm. Sensitivities were 32% and 43%, respectively. Irrespective of contralateral stenosis, a >35% difference in ipsilateral MCA peak systolic velocity relative to the ipsilateral posterior cerebral artery had a 100% specificity and a 23% sensitivity for detecting a <1.5 mm minimal residual lumen diameter.
Conclusions Although the TCD sensitivity for detecting a hemodynamically significant stenosis is relatively low, it can be highly specific (up to 100%). We conclude that TCD enhances the specificity of highly sensitive CDUS criteria for detecting a hemodynamically significant ICA stenosis.
Recent studies have suggested that CEA is beneficial in preventing stroke in symptomatic patients with 70% to 99% stenosis at the origin of the ICA if performed with an acceptable surgical morbidity.1 2 3 CEA in asymptomatic patients with a 60% to 99% stenotic lesion at the origin of the ICA may be beneficial; however, a debate remains as to how its efficacy compares with surgical morbidity at various degrees of stenosis.4 5 6 Natural history studies and the randomized endarterectomy trials of symptomatic carotid stenosis all show a correlation between the incidence of ischemic stroke and the degree of stenosis, particularly when the stenosis is 70% or greater and/or results in decreased pressure distally as measured by oculoplethysmography.1 2 3 7 8 9 Because angiography adds an additional risk to surgery and has its own intrinsic inaccuracies in assessing percent stenosis, including different measurement methods,1 2 we and others have been studying the use of CDUS combined with TCD and MRA for identifying endarterectomy candidates.10 11 12 These methods, however, are best suited for assessment of residual lumen diameter rather than percent stenosis.
The aim of this study was to correlate the PSV obtained by TCD with the minimal residual diameter calculated from the intact CEA specimens removed en bloc and to identify TCD criteria that will increase the specificity of CDUS for detecting a hemodynamically significant stenosis at the origin of the ICA (residual lumen diameter of <1.5 mm).
The choice of a <1.5 mm residual lumen diameter as a hemodynamically significant stenotic lesion is based on two factors. First, our previous study described receiver-operating characteristic curve analyses demonstrating that CDUS can be either a highly sensitive or a highly specific test for identifying such a stenotic lesion.13 Second, our preliminary TCD data suggested that when the residual lumen diameter decreased to 1.5 mm, changes in flow in the ophthalmic artery, carotid siphon (distal ICA), and intracranial circulation occur (ie, it is the point at which pressure and flow change across the stenosis occur).
Here we present 100% specific TCD criteria for identifying a hemodynamically significant stenotic lesion at the origin of the ICA with a residual lumen diameter of <1.5 mm.
Materials and Methods
The current series is a subset of 91 patients consecutively studied with CDUS and TCD.13 Patients had TCD studies performed on average 50 days before CEA (median, 10 days; 90% within 90 days) at the Massachusetts General Hospital Neurosurgical Service between September 1992 and December 1995. Patients were excluded if they had a tandem lesion in the distal ICA determined by MRA and CDUS. The decision to perform CEA was based on CDUS data in conjunction with MRA of the extra- and intracranial arterial circulations and in some cases on conventional selective cerebral angiography to exclude patients with intracranial stenotic lesions. Because contralateral ICA stenosis affects both ipsilateral and contralateral arterial flow patterns, we excluded patients with CDUS-proven contralateral ICA stenosis or occlusion from all analyses comparing the flow in the ipsilateral and contralateral arteries. To calculate the percent difference in PSV between two vessels we used the following formula:
Transcranial Doppler Ultrasonography
TCD was performed in the Vascular Laboratory at the Massachusetts General Hospital using a Medasonics Transpect transcranial Doppler machine and a 2 MHz probe. MCA, ACA, and a PCA PSVs were measured from the transtemporal window and OA, and carotid siphon (distal ICA) flow velocities were measured from the transorbital window. When more than one velocity was recorded for the MCA stem, the highest PSV was used. Our laboratory identifies arteries and defines normal velocities based on published criteria.14 15 16 17 18
CEA specimens were removed en bloc without disturbing the lumen.19 They were promptly placed in formalin and decalcified in nitric acid and EDTA. It had been shown in a previous report that the fixation did not appreciably change the residual lumen diameter of the specimens.20 21 Kodachrome images of each specimen were taken and the length and width of each measured in centimeters. Using a razor blade, the specimens were then sectioned horizontally at 0.2-cm intervals and laid in series so that kodachrome images of each section could be obtained. The Kodachrome image of the section with the smallest residual lumen was scanned by a slide scanner, and the images were stored on floppy diskettes. The lumenal area of each stored image was measured digitally, and the minimal residual diameter was calculated from this area measurement using an Aldus photostyler computer program. The residual lumen diameter measurements were rounded to the nearest 0.1 mm, such that 0.14 mm was recorded as 0.1 mm and 0.15 mm recorded as 0.2 mm.
We assessed percent differences in PSVs in the OA, the carotid siphon, and the MCA, ACA, and PCA stems and the direction of flow in the OA and correlated them with the residual lumen diameter of the most stenotic segment of the endarterectomy specimen. Of the original 91 patients reported with CDUS before endarterectomy in our previous study,13 81 patients (mean age, 69 years; 49 men and 32 women) also underwent transorbital OA TCD evaluation, 49 of whom (mean age, 68 years; 34 men and 15 women) had both transorbital and transtemporal TCD evaluations. We were not able to insonate all vessels in all patients and for the individual comparisons we excluded those with missing data. In addition, those with severe contralateral carotid stenoses were excluded from side-to-side comparisons. Twenty-eight (35%) patients were asymptomatic, 20 (25%) had an ipsilateral hemispheric stroke, 24 (30%) had an ipsilateral hemispheric transient ischemic attack, and 9 (10%) had ipsilateral transient monocular blindness. Among the 49 patients with transtemporal TCD, 12 had severe stenosis or occlusion of the contralateral ICA.
The calculated residual lumen diameter of the 81 endarterectomy specimens included in this study ranged from 0.21 to 2.72 mm, with a mean value of 1.02 mm.
Transorbital (OA/Carotid Siphon) TCD
The OA and the carotid siphon both were evaluated using the transorbital approach. The OA signal was analyzed for direction and velocity of flow. The ipsilateral siphon PSV was compared with the contralateral siphon and expressed as percent difference.
Flow Direction in the OA and Side-to-Side PSV Correlations
OA recordings of the 81 patients were analyzed for the flow direction, PSV, and a comparison of the PSV of the stenotic side to that of the contralateral side. Twenty patients had an ipsilateral reversed OA flow direction (sensitivity 31% and specificity 100%) (Table 1A⇓). OA PSV<20 cm/s was 100% specific (sensitivity 9%). A PSV difference of >50% (ipsilateral OA <contralateral OA) in patients with normal flow direction indicated a residual lumen diameter of <1.5 mm, with a specificity of 87% and a sensitivity of 17%. We observed “hump-like” spectral configuration in the OA pulse wave in patients with a <1.5 mm residual lumen, but this finding was found to be insensitive and was not 100% specific.
Unilateral and Side-to-Side Carotid Siphon PSV Comparisons
In the 22 patients with available data, ipsilateral siphon PSV ranged between 25 and 111 cm/s compared with the contralateral PSVs of 41 to 164 cm/s. A >50% PSV difference between ICA (contralateral>ipsilateral) predicted a residual lumen diameter of <1.5 mm with 100% specificity and a sensitivity of 26% (Table 1B⇑).
Transtemporal TCD allowed comparison of percent difference in PSVs between the ipsilateral and contralateral MCAs, the ipsilateral MCA and PCA, and the ipsilateral and contralateral ACAs.
Side-to-Side MCA PSV Comparisons
In the 45 patients with available data, an MCA PSV <50 cm/s on the ipsilateral side only had a specificity of 71% and a sensitivity of 8% for predicting a residual lumen diameter of <1.5 mm. In the 34 patients with bilateral MCA insonation, PSV on the ipsilateral side ranged from 41 to 128 cm/s and on the contralateral side ranged from 49 to 155 cm/s. A >35% PSV difference between MCAs (contralateral >ipsilateral) predicted a residual lumen diameter of <1.5 mm with 100% specificity and a sensitivity of 32% (Table 2A⇓). The PI of either the ipsilateral MCA or a comparison of ipsilateral/contralateral PIs was not found to be either highly specific or sensitive in predicting a residual lumen of <1.5 mm.
Ipsilateral MCA/PCA PSV Comparisons
In 36 patients with available data, MCA PSVs ranged from 41 to 128 cm/s and PCA PSVs ranged from 36 to 231 cm/s. A >35% PSV difference in ispilateral MCA<PCA predicted a residual lumen diameter of <1.5 mm with 100% specificity and a sensitivity of 23% (Table 2B⇑).
ACA Flow Direction and Side-to-Side PSV Comparison
Of 30 patients with ipsilateral ACA recordings, 10 had reversal of flow (sensitivity 30% and specificity 83%) (Table 2C⇑). One of the 10 patients with a reversal of flow in the ipsilateral ACA had a residual lumen size >1.5 mm (ie, 1.59 mm).
In the 19 patients with available data, the ipsilateral ACA PSVs ranged from 27 to 83 cm/s, and the contralateral PSVs ranged from 45 to 164 cm/s. A >50% PSV difference between contralateral and ipsilateral ACAs predicted a residual lumen diameter of <1.5 mm with a specificity of 100% and a sensitivity of 43% (Table 2D⇑).
Combined 100% Specific TCD Criteria
Criteria with 100% specificity for identifying a <1.5-mm residual lumen diameter are summarized in Table 3⇓. By combining the five most secure of the 100% specific criteria (ipsilateral OA reversal, >50% PSV difference between the carotid siphons, >35% PSV difference between MCAs, >35% PSV difference between ipsilateral PCA and MCA, and >50% PSV difference between ACAs) so that if any one was present, the overall sensitivity of TCD increases to 49% for diagnosing a residual lumen of <1.5 mm (Table 4A⇓). Although an OA PSV <20 cm/s was 100% specific, due to the limited number of observations (n=4), it was not included.
If patients met any of the 100% specific CDUS criteria (a PSV of >440 cm/s, end-diastolic flow velocity of >155 cm/s, or a carotid index of >10) derived from our previous study,13 or if patients met any of the 100% specific TCD criteria derived from this study, a residual lumen diameter of <1.5 mm could be predicted with a sensitivity of 89% (Table 4B⇑).
Of the 49 patients who underwent both CDUS and TCD, 29 patients met the 100% specific CDUS criteria developed in our previous study for a residual lumen diameter <1.5 mm.13 Excluding these patients, and the 1 patient who met neither the highly or moderately specific CDUS criteria, 19 remained who met the highly sensitive (96%) but moderately specific (61%) CDUS criteria for predicting this lesion. Thirteen of these 19 patients actually had a minimal residual lumen diameter <1.5 mm. Ten of these13 patients met one of the five 100% specific TCD criteria. Thus, in these patients, the addition of TCD enhanced the highly sensitive CDUS criteria, increasing the overall specificity to 100%.
These results define the TCD criteria that our laboratory established to detect a stenotic lesion at the origin of the internal carotid artery with a residual lumen diameter of <1.5 mm. One hundred percent specific criteria could only be developed when we used a residual lumen diameter of <1.5 mm. At larger lumen diameters, the hemodynamic changes could not be demonstrated systematically. This suggested that this degree of stenosis was the threshold of hemodynamic significance (ie, the point at which pressure drops across the stenosis and affects flow distally). Using PSV differences and direction of flow data, we identified five criteria with 100% specificity for detecting a residual lumen diameter <1.5 mm.
Individually, each of our five100% specific TCD criteria had relatively low sensitivity; however, if all five criteria are looked for in a given patient but only one is found, the 100% specificity is preserved and the combined sensitivity increases to 49% (Table 4A⇑). These five 100% specific TCD criteria are particularly useful when one of the three 100% CDUS criteria has not been met, thereby leaving in doubt the severity and hemodynamic significance of the stenotic lesion at the origin of the ICA. Combining the three 100% specific CDUS criteria defined in our previous study with the five 100% specific criteria in this study raises the sensitivity of the combined tests to 89% (Table 4B⇑).13
We chose to analyze PSV rather than mean velocity or end-diastolic velocity for two reasons. First, we were concerned that computer-generated mean velocities are inaccurate. End-diastolic velocities have been difficult to measure because the absolute point of end-diastole is obscured by artifact in many cases. PIs are similarly difficult. In contrast, PSVs are easily measured. PIs require manual or computer-generated values for three variables (PSV, EDV, and mean velocity) with their inevitable and compounded measurement errors. Furthermore, using PI we could not find 100% specific criteria for diagnosing a minimal residual lumen diameter of less than or equal to 1.5 mm. In that regard, our data are consistent with those of Wilterdink et al.22 They established their TCD criteria for diagnosing an angiographically derived percent stenosis using Doppler criteria adopted from Langlois et al.23
Second, we and others have observed that absolute PSVs are affected early in the course of a hemodynamically significant stenotic lesion of the internal carotid origin.24 25 26 27 28 But, absolute PSVs are also affected by age, hematocrit, Pco2, cardiac output, and the degree of activation of brain tissue supplied by the artery being insonated.29 30 31 32 33 For these reasons, the percent difference in PSV was used instead of absolute values. PSV generally fell within reported normal ranges.17 18 Six patients had MCA PSVs that fell above the normal range (120 cm/s). Although not systematically reviewed for this paper, all of the patients had preoperative intracranial MRA or conventional angiography to identify those with intracranial stenoses. We were able to review the angiographic findings in all six cases with elevated MCA PSVs and found no evidence of intracranial stenoses. Contralateral ACA or ipsilateral PCA PSV above the reported normal range (110 and 80 cm/s, respectively) may be an indication of collateral flow enhancement due to a hemodynamically significant carotid lesion or, less likely, due to the infrequent occurrence of an intrinsic stenotic lesion in the ACA or PCA.17 PSVs lower than normal in the ipsilateral MCA (<40 cm/s) suggest the presence of a more proximal hemodynamically significant obstructive lesion. In our view, MRA and computerized tomography angiography should be used as adjuncts in confirming the absence of intracranial carotid or proximal middle cerebral artery lesions before endarterectomy.
There are two compelling reasons to identify the TCD criteria that correlate with a hemodynamically significant stenotic lesion at the origin of the ICA. First, the small but definite risk of conventional angiography when measuring percent stenosis has prompted us and others to avoid it in the evaluation of patients with symptomatic and asymptomatic carotid stenoses.10 11 12 We rely instead on CDUS, TCD, and MRA. But, each of these directly or indirectly assesses the actual residual lumen diameter rather than percent stenosis. Thus, it becomes important to correlate the TCD findings of a hemodynamically significant carotid stenosis in terms of residual lumen diameter just as we have done with our CDUS criteria.13
Second, the natural history studies of asymptomatic carotid stenosis7 8 9 and the symptomatic endarterectomy trials1 2 all suggest that the rate of stroke increases with increasing degrees of stenosis, particularly when the stenotic lesion measures 70% or greater using the NASCET angiographic criteria. In addition, NASCET was only able to demonstrate efficacy for carotid endarterectomy in preventing stroke when the degree of stenosis was 70% or greater.1 If one assumes that a lumen diameter of 5 to 6 mm is normal for a distal ICA, then a 1.5 mm residual lumen diameter corresponds to a 70% to 75% stenosis by the NASCET criteria. Our data suggest that this is the point at which a pressure drop across the stenosis occurs in most patients, ie, the point at which the stenotic lesion at the origin of the ICA becomes hemodynamically significant. When this occurs and flow is reduced distal to the ICA stenosis, collateral flow through the OA or circle of Willis is called into play as it is in carotid occlusion.32 33 34 If collateral flow is not adequate, low-flow infarcts or transient ischemic attacks develop, as suggested by Ringelstein et al33 35 and others.36 37 If collateral flow is adequate, patients should remain free of low-flow symptoms.36 But the reduced flow in the ICA may portend thrombus formation at the ICA stenotic site that either occludes it and/or embolizes distally to produce symptoms. We suggest that this is why the natural history studies all show a substantial increase in stroke risk in the asymptomatic patient at 70% or greater stenosis and NASCET could only prove efficacy of surgery at ≥70% stenosis.1 7 8 9 The combined CDUS and TCD data presented here and in our previous study outline highly sensitive and 100% specific criteria for documenting such a lesion.13
We conclude that TCD is a reliable noninvasive test for predicting the presence of a hemodynamically significant extracranial ICA stenosis. If our selected TCD criteria were used, the specificity of noninvasive CDUS could be enhanced.
Selected Abbreviations and Acronyms
|ACA||=||anterior cerebral artery|
|CDUS||=||carotid duplex ultrasonography|
|ICA||=||internal carotid artery|
|MCA||=||middle cerebral artery|
|MRA||=||magnetic resonance angiography|
|NASCET||=||North American Symptomatic Carotid Endarterectomy Trial|
|PCA||=||posterior cerebral artery|
|PSV||=||peak systolic flow velocities|
This paper owes much to the endowment support of Paul O’Neill, the Latsis family, Robert P. Gwinn, Vera and J.W. Gilliland, the Eliot B. Shoolman fund, and the Merrill Lynch Fund for Clinical Research in Cerebrovascular Disease. Drs Can and Suwanwela were John Conway Fellows in cerebrovascular disease at the Massachusetts General Hospital. They now manage stroke services at the Baskent University Hospital in Ankara and the Chulalongkorn Hospital in Bangkok. We acknowledge with appreciation the contribution of the Vascular Laboratory technicians: C. Oliver, RN, RVT; S. Croteau RN, RVT; V. McCallum RN, RVT; D. Henry, RVT; C.X. He, RVT; and D. Moller, RDMS.
- Received January 10, 1997.
- Revision received July 11, 1997.
- Accepted July 21, 1997.
- Copyright © 1997 by American Heart Association
Mayberg MR, Wilson SE, Yatsu F, Weiss DG, Louis M, Hershet LA, Colling C, Eskridge J, Deykin D, Winn HR, for the Veterans Affairs Cooperative Studies Program 309 Trialist Group. Carotid endarterectomy and prevention of cerebral ischemia in symptomatic carotid stenosis. JAMA. 1991;266:3289-3294.
Barnett HJM, Eliasziw M, Meldrum HE, Taylor DW. Do the facts and figures warrant a 10-fold increase in the performance of carotid endarterectomy on asymptomatic patients? Neurology. 1996;46:603-608.
Roederer GO, Langlois YE, Jager KA, Primozich JF, Beach KW, Phillips DJ, Strandness DE Jr. The natural history of carotid arterial disease in asymptomatic patients with cervical bruits. Stroke. 1984;15:603-613.
Kistler JP, Ojemann RG, Crowell RM. Atherothrombotic disease of the carotid circulation: pathophysiology, evaluation and management. In: Ojemann RG, Ogilvy CS, Crowell RM, Heros RC, eds. Surgical Management of Neurovascular Disease. Third ed. Baltimore, Md: Williams & Wilkins; 1995:1-24.
Ogilvy CS, Kistler JP, Ojemann RG, Crowell RM. Carotid endarterectomy without conventional angiography. New England Neurosurgical Society, Fall Meeting, Bretton Woods, NH, 1994.
Suwanwela N, Can U, Furie KL, Southern JF, Macdonald NR, Ogilvy CS, Hansen CJ, Buonanno FS, Abbott WM, Koroshetz WJ, Kistler JP. Carotid Doppler ultrasound criteria for internal carotid artery stenosis based on residual lumen diameter calculated from en bloc carotid endarterectomy specimens. Stroke. 1996;27:1965-1969.
Weschler LR, Ropper AH, Kistler JP. Transcranial Doppler in cerebrovacular disease. Stroke. 1986;17:905-912.
DeWitt LD, Rosengart A, Teal PA. Transcranial Doppler ultrasonography: normal values. In: Babikian VL, Wechsler LR, eds. Transcranial Doppler Ultrasound. St. Louis, Mo: Mosby; 1993:29.
Reutern GM, Budingen HJ. Doppler examination of the basal cerebral arteries: examination of specific cerebral arteries. In: Reutern, GM, Budingen HJ, eds. Ultrasound Diagnosis of Cerebrovascular Disease: Doppler Sonography of the Extra- and Intracranial Arteries Duplex Scanning. New York, NY: Thieme Verlag; 1993:140.
Ojemann RG, Ogilvy CS, Crowell RM, Heros RC. Surgical management of neurovascular disease. In:. Crowell RM, Ogilvy CS, Ojemann RG, Heros RC, eds. Carotid Endarterectomy.Baltimore, Md: Williams & Wilkins; 1995.
Schenk EA, Bond G, Aretz TH, Angelo JN, Choi HY, Rynalski T, Gustafson NF, Berson AS, Ricotta JJ, Goodison MW, Bryan FA, Goldberg BB, Toole JF, O’Leary DH. Multicenter validation study of real-time ultrasonography, arteriography, and pathology: pathologic evaluation of carotid endarterectomy specimens. Stroke. 1988,19:289-296.
Call GK, Abbott WM, Macdonald NR, Megerman J, Davis KR, Heros RC, Kistler JP. Correlation of continuous-wave Doppler spectral flow analysis with gross pathology in carotid stenosis. Stroke. 1988;19:584-588.
Wilterdink JL, Feldmann E, Furie KL, Bragoni M, Benavides JG. Transcranial Doppler ultrasound battery identifies severe internal carotid artery stenosis. Stroke. 1997;28:133-136.
Wechsler LR. Role of transcranial Doppler ultrasonography in clinical practice. In: Babikian VL, Wechsler LR, eds. Transcranial Doppler Ultrasound. St Louis, Mo: Mosby; 1993:305.
Hartl WH, Furst H. Application of transcranial Doppler sonography to evaluate cerebral hemodynamics in carotid artery disease. Stroke. 1995;26:2293-2297.
Schneider PA, Rossman ME, Bernstein EF, Ringelstein EB, Otis SM. Noninvasive assessment of cerebral blood supply through the ophthalmic artery. Stroke. 1991;22:31-36.
Schneider PA, Rossman ME, Bernstein EF, Torem S, Ringelstein EB, Otis SM. Effect of internal carotid artery occlusion on intracranial hemodynamics. Stroke. 1989;19:589-593.
Lassen NA, Astrup J. Cerebral blood flow: normal regulation and ischemic thresholds. In: Weinstein PR, Faden AI, eds. Protection of the Brain From Ischemia. Baltimore, Md: Williams &Wilkins; 1990.
Tranmer BI, Keller TS, Kindt GW, Archer D. Loss of cerebral autoregulation during cardiac output variations in focal cerebral ischemia. J Neurosurg. 199277:253-259.
Bragoni M, Feldmann E. Transcranial Doppler indices of intracranial hemodynamics. In: Tegler CH, Babikian VL, Gomez CR, eds. Neurosonology. Baltimore, Md: Mosby; 1996.
Ringelstein EB, Sievers C, Ecker S, Schneider PA, Otis SM. Noninvasive assessment of CO2-induced cerebral vasomotor response in normal individuals and patients with internal carotid artery stenosis. Stroke. 1988;19:963-969.
Schenider PA, Rossman ME, Bernstein EF, Torem S, Ringelstein BE, Otis SM. Effect of internal carotid artery occlusion on intracranial hemodynamics: transcranial Doppler evaluation and clinical correlation. Stroke. 1988;19:589-593.
Ringelstein BE, Zeumer H, Angelou D. The pathogenesis of strokes from internal carotid artery occlusion: diagnostic and therapeutic implications. Stroke. 1983;14:867-875.
Mull M, Schwarz M, Thron A. Cerebral hemispheric low-flow infarcts in arterial occlusive disease: lesion patterns and angiomorphological conditions. Stroke. 1997;28:118-123.