Eur J Cardiothorac Surg 2007;32:521-526. doi:10.1016/j.ejcts.2007.05.024
Copyright © 2007, European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved
Value of cerebral perfusion computed tomography in the management of intensive care unit patients with suspected ischaemic cerebral pathology after cardiac surgery
Sotirios Bisdasa,*,
Panagiotis Therapidisb,
Josef Matthias Kerla,
Nestoras Papadopoulosb,
Iris Burcka,
Christopher Herzoga,
Thomas J. Vogla
a Department of Radiology, Johann Wolfgang Goethe University Hospital, Frankfurt, Germany
b Department of Cardiothoracic Surgery, Johann Wolfgang Goethe University Hospital, Frankfurt, Germany
Received 27 December 2006;
received in revised form 30 May 2007;
accepted 31 May 2007.
* Corresponding author. Address: Department of Radiology, Johann Wolfgang Goethe University Hospital, Theodor Stern Kai 7, D-60590 Frankfurt, Germany. Tel.: +49 69 6301 7260; fax: +49 69 6301 5252. (Email: s.bisdas{at}med.uni-frankfurt.de).
 |
Abstract
|
|---|
Objective: Adverse neurologic outcomes, like stroke, in intensive care unit (ICU) patients after cardiac surgery can have devastating consequences, among them increased mortality risk and, among survivors, loss of independence and a diminished quality of life. Non-contrast computed tomography (CT) remains a widely utilised modality for assessing stroke; however, it has a low sensitivity in the acute phase. Perfusion CT (PCT) has the potential of imaging stroke in its hyperacute phase. We evaluated the feasibility and results of the method among patients from the ICU. Methods: The NCCT and PCT images of 33 retrospectively identified patients were included in this study. The diagnostic contribution of the PCT to patient management was classified according to one of three categories: (A) those that changed the preliminary (NCCT) diagnosis; (B) those that revealed additional pathology and/or specified more exactly findings that have been detected by NCCT or clinically suspected; and (C) confirmed the preliminary diagnosis. Neurologic outcome variables were also documented and associated with PCT lesions. Results: Fifteen patients after coronary artery bypass graft (CABG) operation, 14 patients after CABG and valve surgery, and 4 patients after an aortic dissection (Type A) surgery underwent a NCCT with PCT 2.4 ± 1.3 days after the operation. Twenty patients had bilateral internal carotid artery (ICA) stenosis (>50%), 11 patients had unilateral ICA stenosis (>75%), and 2 patients had no ICA stenosis. In nine patients (27.2%) the PCT changed the initial diagnosis of the NCCT and revealed ischaemic pathology. In 24 patients (72.7%), the performed PCT revealed additional pathology and/or more completely characterised findings that have been detected by the initial NCCT. In nine patients, PCT confirmed only the initial diagnosis. Patients with normal PCT findings had a favourable outcome; patients with large lesions in PCT in one or more vascular territories had an unfavourable outcome; seven patients with lesions in basal ganglia and/or semioval centre had a favourable outcome. Conclusions: PCT shows a greater sensitivity in detecting and mapping acute ischaemic stroke in ICU patients (after cardiac surgery) in whom conventional imaging findings are not in line with the severity of the clinical condition.
Key Words: CT • Perfusion CT • Intensive care unit • Stroke
|
1. Introduction
|
|---|
Despite an overall decrease in perioperative and postoperative morbidities and mortalities, evidence of some degree of central nervous system dysfunction associated with cardiac surgery—with or without cardiopulmonary bypass—has steadily increased. Patient-specific factors (like pre-existing extracranial or intracranial atherosclerotic disease, haematologic mechanisms such as blood-clotting mechanisms) have a fundamental impact on the risks of a cerebral embolisation and/or ischaemic hypoperfusion that result in a perioperative brain injury [1].
Non-contrast CT (NCCT) imaging of the head is still the most widely accepted imaging modality in the world for the emergency setting of an acute stroke. However, the sensitivity of the NCCT in identifying the ischaemic regions and predicting the infarct extent which correlate with the clinical outcome is not high [2]. Perfusion CT (PCT) helps in this respect by allowing non-invasively the judgment of the brain parenchyma affection by means of quantitative perfusion values, including cerebral blood flow (CBF), cerebral blood volume (CBV), and mean transit time (MTT). Several studies have shown that PCT correlates well with other vascular imaging modalities [3]. The entire investigation requires only a few minutes and provides a similar amount of information as stroke MRI helps defining the infarcted region, the penumbra, and predicting the infarct extension, at least in a short follow-up [4].
Although PCT is widely accepted for the evaluation of acute stroke patients admitted to the emergency room, there is no data about the feasibility of PCT imaging and the advantages, which may offer, in the assessment of the intensive care unit (ICU) patients, who have undergone cardiac surgery, and present with an unknown or unclear cerebral pathology underlying their serious clinical condition. We sought to address this influence of PCT by determining the confirmation or alteration of the preliminary/working diagnosis (based on NCCT findings).
|
2. Materials and methods
|
|---|
We retrospectively identified 218 patients from the intensive care cardiovascular unit, seen between September 2004 and December 2006, who presented symptoms of an acute stroke after a cardiovascular operation and had received an NCCT in order to exclude stroke. In 33 patients, there was a disparity between their NCCT findings and their clinical status and an additional PCT was performed. The NCCT and PCT images of these 33 patients were included in this study. In this patient population, there was no patient with renal insufficiency, hyperthyroidism, or known allergy to iodinated contrast agent. In the case of patients with pre-existing neurological disease or previous stroke that would hamper interpretation of imaging data, we included their data only if we had a preoperative NCCT in order to identify reliably the possible new cerebral lesions.
2.1 Imaging protocol
Standard baseline and follow-up NCCT scanning was performed in a 16-row multislice unit (Sensation 16, Siemens, Erlangen, Germany) using the following parameters: 120 kV, 180 mA, 512 x 512 image matrix, a 22-cm DFOV (Displayed Field of View) and 5 mm (posterior fossa/infratentorial)/10 mm (supratentorial) slice thickness.
PCT consisted of a 45-s series performed in four 6-mm thick adjacent brain sections during the intravenous administration (45 ml) of a 400 mg/dl iodinated non-ionic contrast material (Imeron 400, Altana, Germany). The most caudal section of the perfusion CT series was at the slice of the NCCT at the level of the third ventricle and the basal ganglia, positioned above the orbits to protect the lenses. There was no movement of the patient in the time interval between NCCT and PCT and the gantry tilt was the same.
The contrast agent was administered into an antecubital vein by using a power injector (Medrad, Indianola, PA) at an injection rate of 6 ml/s. The acquisition parameters were 80 kVp and 120 mAs. CT scanning was initiated 6 s after the start of the injection. The time delay before contrast material reached the brain parenchyma allowed the acquisition of non-enhanced baseline images, required for the post-processing of the perfusion CT data.
2.2 Data analysis
The patient data was blinded and two radiologists performed the readings in consensus and were blinded to the results of the clinical examinations. The NCCT studies (including those obtained preoperatively) were evaluated for conventional signs of stroke: hemispheric hypoattenuation or swelling, loss of the cortical ribbon (including the insular ribbon), hypoattenuation in the basal ganglia, and the hyperdense artery sign. Furthermore, the reviewers assessed hypoattenuation in the arterial territories. If a hypoattenuated brain area was found, the extent of the hypoattenuation was documented.
PCT data were analysed using commercially available PCT software (TeraRecon Inc., USA) based on the central volume principle, which is the most accurate for low injection rates of iodinated contrast material. Motion correction was not necessary for any patient. The applied deconvolution operation requires a reference arterial input function (AIF), which was selected by the software in a region of interest (ROI) that the user placed in the anterior cerebral artery (ACA). The venous outflow function ROI was placed in the superior sagittal sinus. CBF, CBV and MTT colour maps were generated by the software. The same two radiologists, with experience in PCT imaging, evaluated in consensus the coloured perfusion maps for abnormalities. Firstly, the readers evaluated the MTT map for time delays of the arrival of the contrast agent that are indicative of an arterial occlusion. The MTT maps are considered to be most sensitive in detecting acute stroke. Secondly, the CBV maps were evaluated for any abnormalities, which were correlated if possible with the MTT findings. Finally, the CBF maps were observed for any hypoperfused or hyperperfused areas that indicated ischaemic or luxury hyperperfused areas, respectively. Special care was taken not to include cerebrospinal fluid (CSF), or large vessels in the defined ROI. In case of unclear findings in PCT maps, the ROI with the suspected lesion was mirrored on the contralateral hemisphere in order to compare the perfusion values with the corresponding presumably healthy regions. If any perfusion value in the suspected lesion was different from the corresponding value in the healthy hemisphere for at least 2 SD, then it was considered pathologic.
The diagnostic contribution of the PCT to patient management was established by one radiologist and one cardiovascular surgeon in consensus, and was classified according to three categories:
- • Those that changed the preliminary (NCCT) diagnosis.
- • Those that revealed additional pathology and/or specified more exactly findings that had been detected by CT or clinically suspected.
- • Those that confirmed the preliminary (NCCT) diagnosis.
As the goal of our study was to demonstrate the usefulness of PCT imaging in cardiovascular ICU patients, no correlation analysis between location and/or size of the lesions and clinical scores was attempted. The outcome diagnosis in our patients was defined as Type I neurologic outcomes, which included stroke, transient ischaemic attack (TIA), coma at discharge, or death due to stroke or hypoxic encephalopathy or Type II neurologic deficits such as deterioration in intellectual function, confusion, agitation, disorientation, memory deficit or seizure. The reporting period included the time from the admission in the ICU to discharge from the ICU or from the hospital.
|
3. Results
|
|---|
Nineteen males and 14 females were included in the PCT study. The mean age of the patients was 69.1 ± 5.6 (SD) years. Fifteen patients were admitted in ICU after coronary artery bypass graft (CABG) operation, 14 patients after CABG and valve surgery and 4 patients after aortic dissection (Type A) surgery. Among the CABG patients, the vast majority (82%) underwent an On-pump surgery. Twenty patients had bilateral internal carotid artery (ICA) stenosis (>50%), 11 patients had unilateral ICA stenosis (>75%) and 2 patients had no ICA stenosis. The patients underwent an NCCT with PCT 2.4 ± 1.3 (SD) days after the operation. The NCCT and PCT findings of the patient population are shown in Table 1
. In nine patients (27.2%), PCT changed the initial diagnosis of the NCCT and revealed ischaemic pathology. In 24 patients (72.7%), the performed PCT revealed additional pathology and/or more completely characterised findings that have been detected by the initial NCCT. Finally, nine patients did not apparently profit from the subsequent PCT, as the latter did not change the initial diagnosis. Regarding the neurologic outcome of our subjects, patients with normal findings in PCT had a favourable outcome; patients with large lesions in PCT in one vascular territory or with lesions in more than one vascular territories were classified as Type I patients; seven patients with lesions (PCT) in basal ganglia and/or semioval centre had a Type II neurologic outcome. Two illustrative cases showing the NCCT and PCT of patients who benefited by the PCT imaging are shown in Figs. 1 and 2
.
View larger version (118K):
[in this window]
[in a new window]
|
Fig. 1. (A–D) Non-contrast CT (NCCT) scan and perfusion CT (PCT) of a 65-year-old male patient after coronary artery bypass graft with valve surgery. The patient developed a left-sided hemiparesis after 3 days in the intensive care unit. In this case, there was a disparity between initial CT imaging and clinical condition, and, thus, a PCT was performed. PCT changed the initial diagnosis and revealed a right-sided ischaemia. (A) The NCCT demonstrates no signs of an acute ischaemia. Note only the periventricular hypoattenuation which indicates a chronic vasculopathy. (B) Cerebral blood flow (CBF) parametric map of PCT imaging shows reduced CBF in the grey and white matter in the anterior, middle and posterior cerebral artery territories. (C) In the cerebral blood volume (CBV) parametric map, there is no demarcation of the ischaemic regions, probably due to the reactive vasodilatation in the early phase of the stroke. (D) The mean transit time (MTT) map clearly shows the regions with higher MTT of the contrast agent (green colour) which correspond to the regions with low CBF.
|
|
View larger version (104K):
[in this window]
[in a new window]
|
Fig. 2. (A–D) Non-contrast CT (NCCT) scan and perfusion CT (PCT) of a 71-year-old male patient after coronary artery bypass graft with valve surgery. The patient developed a severe left-sided hemiparesis after 2 days in the intensive care unit. A disparity between initial NCCT and clinical condition was noted, and a PCT was subsequently performed. This case illustrates the additional benefits of PCT in the exact localisation of the ischaemia without changing the initial diagnosis. (A) NCCT demonstrates a sulcal effacement in the middle cerebral artery territory (arrows) which is an early sign of acute ischaemia. (B) Cerebral blood flow (CBF) parametric map of the PCT imaging shows, however, reduced CBF in the grey and white matter in the anterior and posterior cerebral artery territories and in the basal ganglia, too. (C) In the cerebral blood volume (CBV) parametric map, there is no demarcation of the ischaemic regions, probably due to the autoregulatory vasodilatation. (D) The mean transit time (MTT) map clearly shows the regions (in the whole right hemisphere) with higher MTT of the contrast agent (green colour) which correspond to the regions with a delay of the contrast agent arrival and to those with low CBF.
|
|
|
4. Discussion
|
|---|
The sensitivity of NCCT to infarction within the first hours after stroke onset is low, and ischaemic signs are not readily recognised. Previous studies [5–7] have revealed a moderate interobserver reliability in the detection of early CT hypoattenuation, which represents the early phase of acute infarction corresponding to the cytotoxic edema. This stage is followed by the vasogenic edema and a conspicuous hypoattenuation, which is followed by the ‘fogging effect’ in the subacute stage. The early signs of infarction deliver no quantitative information about the irreversibly infarcted area or the potentially salvageable penumbra zone. Section-selective PCT involves the acquisition of sequential CT data from a few sections in a cine mode during a small (40–45 ml) intravenous bolus of iodinated contrast material. PCT data can be analysed with commercially available software according to two distinct mathematical models: the maximal slope model and the central volume principle. PCT imaging overcomes the shortcomings of NCCT and offers a rapid and reliable assessment of cerebral haemodynamics in patients with cerebrovascular injury. Especially for the diagnostic imaging of ICU patients, it is important that the selected imaging modality offers rapid evaluation in the radiology department. PCT can provide more accurate evaluation of the cerebral haemodynamics, compared to other CT imaging methods like CT angiography, and our PCT protocol involves the least possible radiation dose required for a non-invasive assessment of the brain haemodynamics. Compared to MR imaging and perfusion-weighted MR (PWI), PCT has the major advantage of being able to assess all cerebral perfusion parameters in a robust, quantitative way, and thus allows a direct insight into cerebral vascular autoregulation and can describe the salvageable penumbra. PWI lacks absolute quantitative measurements and, therefore, can describe penumbra as the difference between the PWI and diffusion-weighted (DWI) abnormalities. It should be emphasised that the PWI is dependent on acquisition sequence factors (e.g. echoplanar imaging or gradient-echo sequences), contrast agent dosage and MR field strength, while MR units are not so widely available as CT scanners [9]. Finally, the data post-processing of PCT is easier concerning the AIF sampling compared to the PWI. DWI is also very useful in detecting acute ischaemia and previous study has shown that the diffusion abnormalities correlate excellently with the PCT abnormalities in hyperacute brain ischaemia [4]. However, its availability is limited, too. Finally, the serious clinical condition, which necessitates respiratory support and continuous monitoring, and the longer duration of the MR surveys are further drawbacks of MR examinations in ICU patients.
The limited spatial coverage of dynamic PCT might represent a major drawback and jeopardise its sensitivity. With multidetector-row CT technology used in our study, the spatial coverage of dynamic PCT can reach 20–24 mm for each bolus of contrast material. When we consider the amount of the contrast agent and the risk of nephrotoxicity, a maximum of two 40–50-ml boluses can be used; therefore, total coverage up to 40–48 mm can be reached. Furthermore, modern CT units can cover 40 mm in one slab and two successive injections can cover the entire hemispheres. To increase coverage with PCT during a single bolus of contrast agent, certain technical manipulations, such as the toggling-table technique, have been proposed [8]. The limitation of PCT in coverage can also be efficiently solved if the examined level is through the basal ganglia. A previous study reported a sensitivity of 95% in the detection of territorial perfusion deficits [5]. In our study, detection of the ischaemic region was possible in all patients. Nevertheless, calculation of the entire volume of ischaemia remains limited in cases in which the lesion extends beyond the covered area. The injection rate of 6 ml/s, which was used in our study in order to acquire a bolus arrival of the contrast agent, may be high and, thus, may be associated with some risk of contrast leakage. We think that the software calculations used in this study may be achieved with lower injection rates, i.e. 4 ml/s, compared to the maximum slope model analysis which requires very high injection rates [10].
The utilisation and possible benefits of PCT in patients of the ICU after cardiac surgery has not yet been investigated. This group of patients suffers from an increased risk for a postoperative ischaemic brain injury which may increase the hospital stay as well as deteriorate the long-term outcome. The incidence of clinically obvious strokes after CABG is reported to be between 0.8 and 5.2% [11]. The typically poor postoperative course of patients who develop stroke after cardiac surgery underlines the need for timely recognition, prevention/modification of factors that predispose to stroke and it implies certain clinical, ethical and socio-economic issues. In our study, findings on PCT studies contributed significant information in 24 of 33 (72%) patients. The PCT findings revealed acute infarction (contrary to the NCCT findings) and, thus, changed the initial diagnosis in nine patients. This group of patients benefited mostly from PCT, while in 10 other patients PCT revealed more lesions than NCCT. Though, it may be assumed that these cerebrovascular lesions would remain silent, it is possible that some of them were responsible for the disparity between the NCCT findings and clinical status. Finally, in five patients PCT demonstrated the real extent of abnormality, which was different from that shown in NCCT. The agreement between normal NCCT and PCT findings in nine patients, despite their clinical condition, does not necessarily show a pitfall of the PCT imaging. The occurrence of clinically obvious stroke, which was the outcome measure of our studies, is not easily identified. Thus, postoperative cognitive impairment may or may not be classified as ‘stroke’, and these patients may be classified in other categories, such as delirium, depression and dementia, or vice versa. Therefore, the patients referred for imaging may not reflect the true incidence of postoperative strokes.
Embolic phenomena have been previously implicated in the pathophysiology of stroke after On-pump CABG, whereas myocardial stunning and hypoperfusion (resulting in transient ischaemic attacks (TIAs)) may be possible mechanisms associated with delayed onset of stroke after Off-pump CABG [12]. It is not assured that multiple territory cerebral microinfarcts (as a result of extended embolic phenomena) or TIAs can be reliably identified with PCT. This implies that both the NCCT and PCT delivered false-negative results and their agreement is misleading. At this field, the combined use of DWI and PW may provide better results [13], though prospective studies are required and the clinical implications of TIAs imaging are not ‘evidence-based’ [14].
Some sceptics may argue for the therapeutic implications of the application of PCT as a straightforward and sensitive diagnostic tool for detecting cerebral ischaemia in patients after cardiac surgery. Seen from a descriptive point of view, the involvement of one or more large ischaemic territories in PCT was associated with an unfavourable outcome, while an isolated lesion in one more site seemed to result in Type II neurologic outcome. Although an aggressive lysis of the causative thrombus may be an inappropriate therapeutic option, basic acute management, including normalisation of blood glucose and body temperature, volume therapy, maintaining a high blood pressure and cardiac output to improve remaining cerebral perfusion in the presence of ischaemically impaired autoregulation, treating cerebral oedema, prophylaxis of thrombosis, and early mobilisation, arrhythmia therapy as well as management of contributory factors, like systemic inflammatory response, may in the early stage minimise the existing cerebral injury and avoid a new one.
The value of the use of preoperative characteristics to identify patients who are at greatest risk is the potential to alter care and give appropriate information to clinicians and patients in their treatment decisions. A stroke risk index including PCT, on the basis of preoperative characteristics may be developed to predict a patient's risk of perioperative stroke. PCT may identify certain patients with abnormal flow patterns and impaired cerebrovascular reserve capacity. In such cases, PCT maps before and after vasodilatator stimulation may be quantified and a side-to-side comparison of the cerebral hemispheres may reveal impaired vascular supply or vascular reactivity; thus, a modification of intraoperative algorithms, e.g. temperature or blood pressure management, may prevent injury [15].
Certain drawbacks of this study are the small patient population, the absence of a quantification of the perfusion parameters and their correlation with the clinical scores for admission as well as discharge. Thus, the known predictive role of PCT in the long-term outcome was not be verified in our patient population [4]. Nevertheless, we think that these drawbacks do not understate our findings which aim to highlight the primary diagnostic advantages of PCT in ICU patients.
In conclusion, the results of this work demonstrate the potential role of PCT in ICU patients (after cardiac surgery) in whom conventional imaging findings are not in line with the severity of the clinical condition. PCT has shown a greater sensitivity in detecting and mapping acute ischaemic stroke, and the implementation of PCT in the clinical routine CT surveys for ICU patients may be beneficial.
|
References
|
|---|
- Hedberg M, Boivie P, Edstrom C, Engstrom KG. Cerebrovascular accidents after cardiac surgery: an analysis of CT scans in relation to clinical symptoms. Scand Cardiovasc J 2005;39(5):299-305.[CrossRef][Medline]
- Parsons MW, Pepper EM, Chan V, Siddique S, Rajaratnam S, Bateman GA, Levi CR. Perfusion computed tomography: prediction of final infarct extent and stroke outcome. Ann Neurol 2005;58:672-679.[CrossRef][Medline]
- Kloska SP, Nabavi DG, Gaus C, Nam EM, Klotz E, Ringelstein EB, Heindel W. Acute stroke assessment with CT: do we need multimodal evaluation?. Radiology 2004;233:79-86.[Abstract/Free Full Text]
- Bisdas S, Donnerstag F, Ahl B, Bohrer I, Weissenborn K, Becker H. Comparison of perfusion computed tomography with diffusion-weighted magnetic resonance imaging in hyperacute ischemic stroke. J Comput Assist Tomogr 2004;28:747-755.[CrossRef][Medline]
- Reichenbach JR, Röther J, Jonetz-Mentzel L, Herzau M, Fiala A, Weiller C, Kaiser WA. Acute stroke evaluated by time-to-peak mapping during initial and early follow-up perfusion CT studies. AJNR Am J Neuroradiol 1999;20:1842-1850.[Abstract/Free Full Text]
- Marks MP, Holmgren EB, Fox AJ, Patel S, von Kummer R, Froehlich J. Evaluation of early computed tomographic findings in acute ischemic stroke. Stroke 1999;30:389-392.[Abstract/Free Full Text]
- Wintermark M, Fischbein NJ, Smith WS, Ko NU, Quist M, Dillon WP. Accuracy of dynamic perfusion CT with deconvolution in detecting acute hemispheric stroke. Am J Neuroradiol 2005;26:104-112.[Abstract/Free Full Text]
- Roberts HC, Roberts TP, Smith WS, Lee TJ, Fischbein NJ, Dillon WP. Multisection dynamic CT perfusion for acute cerebral ischemia: the "toggling-table" technique. Am J Neuroradiol 2001;22:1077-1080.[Abstract/Free Full Text]
- Essig M, Lodemann KP, Bruning R, Kirchin M, Reith W. Intraindividual comparison of gadobenate dimeglumine and gadobutrol for cerebral magnetic resonance perfusion imaging at 1.5T. Invest Radiol 2006;41:256-263.[CrossRef][Medline]
- Hu CH, Wu QD, Hu XY, Fang XM, Zhang TH, Ding Y. Hemodynamic studies on brain CT perfusion imaging with varied injection rates. Clin Imaging 2007;31:151-154.[CrossRef][Medline]
- Stamou SC, Hill PC, Dangas G, Pfister AJ, Boyce SW, Dullum MK, Bafi AS, Corso PJ. Stroke after coronary artery bypass: incidence, predictors, and clinical outcome. Stroke 2001;32:1508-1513.[Abstract/Free Full Text]
- Peel GK, Stamou SC, Dullum MK, Hill PC, Jablonski KA, Bafi AS, Boyce SW, Petro KR, Corso PJ. Chronologic distribution of stroke after minimally invasive versus conventional coronary artery bypass. J Am Coll Cardiol 2004;43:752-756.[Abstract/Free Full Text]
- Restrepo L, Jacobs MA, Barker PB, Wityk RJ. Assessment of transient ischemic attack with diffusion and perfusion-weighted imaging. AJNR Am J Neuroradiol 2004;25:1645-1652.[Medline]
- Laloux P, Jamart J, Meurisse H, Coster P, Laterre C. Persisting perfusion defect in transient ischemic attacks: a new clinically useful subgroup?. Stroke 1996;27:425-430.[Abstract/Free Full Text]
- Bisdas S, Nemitz O, Berding G, Weissenborn K, Ahl B, Becker H, Donnerstag F. Correlative assessment of cerebral blood flow obtained with perfusion CT and positron emission tomography in symptomatic stenotic carotid disease. Eur Radiol 2006;16:2220-2228.[CrossRef][Medline]
Top
Abstract
1. Introduction
