| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Nuclear Medicine |
1 From the Section of Neuroradiology (C.P.D.) and Div of Radiological Sciences (C.P.D., T.O.V., N.R.S., R.L.G., W.J.P.), Mallinckrodt Institute; Dept of Neurology and Neurological Surgery (T.O.V., K.D.Y., S.M.F., R.L.G., W.J.P.); and Lillian Strauss Institute of the Jewish Hospital of St Louis (W.J.P.), Washington University School of Medicine, 510 S Kingshighway Blvd, St Louis, MO 63110. From the 1998 RSNA scientific assembly. Received Sep 16, 1998; revision requested Nov 17; revision received Dec 7; accepted Mar 2, 1999. Supported by National Institutes of Health grants NS02029 (C.P.D.), NS28947 (T.O.V., S.M.F., R.L.G., W.J.P.), and NS34050 (W.J.P.) and Charles A. Dana Foundation (C.P.D., K.D.Y., W.J.P.) and by RSNA Research and Education Foundation as a 1994 Siemens Medical Systems/RSNA Fellow (C.P.D.). Address reprint requests to C.P.D. (e-mail: derdeyn@mirlink.wustl.edu).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: The outcome analysis of a blinded prospective study designed to determine if increased OEF was an independent predictor of stroke in patients with symptoms and with carotid occlusion was repeated by substituting a count-based method of OEF measurement for the original quantitative technique. The performance of the quantitative and count-based methods was assessed by using Kaplan-Meier cumulative survival functions (log-rank, [P < .05]). Receiver operating characteristic (ROC) curves for both methods were generated.
RESULTS: Thirteen ipsilateral strokes occurred during a mean follow-up of 3.1 years for 81 patients. All ipsilateral strokes occurred in 50 patients with increased count-based OEF (P = .002, sensitivity 100%, specificity 46%). Sixty-eight patients underwent complete quantitative studies, which allowed comparison of OEF methods. Both the count-based and the quantitative methods were predictive of stroke in this subgroup (P = .005 and .025, respectively). ROC analysis demonstrated a greater area under the curve for the count-based OEF method.
CONCLUSION: Count-based PET measurement of OEF without arterial sampling accurately predicts stroke in patients with carotid occlusion.
Index terms: Brain, emission CT (ECT), 10.12163 • Brain, ischemia, 10.781 • Carotid arteries, stenosis or obstruction, 172.721, 175.721 • Emission CT (ECT), experimental studies, 10.12163
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
With more severe reductions in cerebral perfusion pressure, the capacity for compensatory vasodilation is exceeded, and cerebral blood flow begins to decline (Fig 1). A progressive increase in OEF then maintains cerebral oxygen metabolism and brain function (stage 2, Fig 1). This most severe form of cerebral hemodynamic failure also has been termed "misery perfusion" (2). At present, OEF can be measured only with PET and oxygen 15–labeled radiotracers.
|
Findings of the St Louis Carotid Occlusion Study (STLCOS) recently demonstrated that increased OEF was an independent predictor of subsequent stroke in patients with symptomatic carotid occlusion in a blinded longitudinal study (12). In PET studies in patients with carotid occlusion before and after extracranial-to-intracranial arterial bypass, the OEF elevation has been shown to be reversible (2,13–15). Consequently, a clinical trial of surgical revascularization for patients with increased OEF has been proposed (12).
The quantitative measurement of an in vivo physiologic process, such as cerebral OEF, with PET is complicated (16). Three basic components are necessary: (a) a radiotracer, (b) a radiation detection system, and (c) a mathematic model relating the physiologic process to the detected radiation. For the quantitative measurement of OEF, a subject first inhales trace amounts of 15O-labeled oxygen mixed with air. The 15O-labeled oxygen is then bound to hemoglobin and transported throughout the body where some of it is used for oxidative metabolism. Some 15O-labeled oxygen is metabolized to water, which returns to the bloodstream and recirculates.
As these processes are occurring, the 15O radionuclides are decaying by emission of a positron. These positively charged particles can travel a few millimeters within a tissue before encountering an electron. This interaction leads to annihilation of both the positron and the electron and generates two photons of equal energy heading in opposite directions. The PET scanner uses data from detector pairs to create an image of regional activity, or PET counts, and this PET image of regional counts must be processed to translate regional activity to a physiologic measurement.
The quantitative measurement of OEF requires time-activity data acquired from arterial sampling during scanning and quantitative values of blood flow and blood volume provided from separate PET examinations (17). This information is used to correct for the PET counts measured in the head owing to unextracted 15O-labeled oxygen, which remains bound to hemoglobin, and the PET counts from recirculating 15O-labeled water owing to metabolism of 15O-labeled oxygen.
Metabolic processing of PET images is complicated and time-consuming. Percutaneous arterial cannulation may not be possible in some patients. Arterial sampling requires additional equipment and adds time to the PET procedure. In the STLCOS, complete arterial time-activity data were not available owing to technical factors in 13 of 81 subjects. These 13 subjects were therefore assigned to categories on the basis of the count-based method of OEF measurement presented in this report, while the remaining 68 subjects were assigned to categories by using the quantitative OEF processing. A count-based technique that does not require arterial time-activity data would be useful, particularly in the setting of a multiinstitutional trial in which standardization of complicated quantitative PET techniques might be difficult.
In this article, we present a further evaluation of the count-based PET technique to measure relative OEF (Fig 2). The count-based OEF image is generated by dividing the PET counts in an 15O-labeled oxygen scan by the PET counts in an 15O-labeled water scan on a pixel-by-pixel basis. The ratio of the counts in the 15O-labeled oxygen image and the 15O-labeled water image are linearly proportional to quantitative regional OEF except for the small contributions from intravascular oxygen and recirculating labeled water. The errors are small when regional oxygen metabolism is normal (17–19).
|
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Patients were followed-up by the study coordinator (S.M.F.) for the duration of the study by telephone contact every 6 months with the patient or next of kin. The occurrence of any symptoms suggesting a stroke was thoroughly evaluated by one designated investigator (R.L.G.), who remained blinded to the PET data. All living patients were followed-up for the duration of the study. The primary end point of the original prospective study was subsequent ischemic stroke defined clinically as a neurologic deficit of presumed ischemic cerebrovascular cause lasting more than 24 hours in any cerebrovascular territory. Secondary end points were ipsilateral ischemic stroke and death.
Eighteen healthy control subjects (eight women, 10 men; age range, 19–77 years; mean age ± SD, 45 years ± 18) were recruited by means of public advertisement. All underwent neurologic evaluation, magnetic resonance (MR) imaging of the head, and duplex ultrasonographic (US) imaging of the extracranial carotid arteries. None had (a) signs or symptoms of neurologic disease other than mild distal sensory loss in the legs consistent with age, (b) pathologic lesions on MR images (mild atrophy and punctate asymptomatic white matter abnormalities were not considered pathologic), or (c) more than 50% stenosis of the extracranial carotid arteries.
A computed tomographic (CT) scan of the brain was obtained without contrast enhancement if CT or MR imaging had not been performed as part of the usual clinical care sufficiently long after an ischemic event to permit accurate definition of infarction location. This CT scan was used to determine the site of tissue infarction to exclude these regions from subsequent PET analysis.
This research was approved by the human studies committee and written informed consent was obtained from all subjects.
PET Examination
Hemodynamic PET studies in patients with carotid occlusion were performed at study entry. Blood pressure was measured in the clinic prior to walking to the scanner suite. After positioning the patient in the scanner gantry, an individually molded thermoplastic face mask was applied to ensure that the patient's head remained in a constant position during the scanning period. The exact position of the patient's head relative to the scanning plane was recorded on a lateral skull radiograph obtained after head immobilization. Venous and, when possible, arterial catheters were placed for the intravenous administration of the radiotracer and for arterial blood gas analyses and arterial time-activity curve determination, respectively (21). All PET studies were performed with use of one of two scanners (ECAT 953B or ECAT EACT HR; Siemens/CTI, Knoxville, Tenn). A transmission scan was obtained by using the two-dimensional mode (intersection septa extended) before radiotracer administration by using germanium 68/gallium 68 rotating rod sources.
Each PET study consisted of three separate physiologic studies. During each, arterial blood samples were drawn by hand or automatically to convert quantitative regional radioactivity data to quantitative physiologic measurements. Additional arterial samples were drawn at intervals during the examination for determination of PaCO2 stability, mean arterial oxygen content calculations, and carboxyhemoglobin content.
A 5-minute scan was obtained 1 minute and 45 seconds after inhalation of one to two breaths of air containing trace amounts of 15O-labeled carbon monoxide. After allowing this activity to decay for approximately 15 minutes, a 40-second scan was obtained after inhalation of one to two breaths of air containing trace amounts of 15O-labeled oxygen. After another 15 minutes, the last scan was acquired after the injection of a bolus of 15O-labeled water.
Both the 15O-labeled oxygen scans and 15O-labeled water scans acquired counts for 40 seconds, beginning when the radiotracer reached the brain. The circulation time for intravenously administered or inhaled radiotracer to reach the brain varies greatly from person to person, so accurate online monitoring of brain radioactivity is necessary to start data acquisition at the proper time. Both scanners used in this study provide such online monitoring of coincidence events within the field of view, but the temporal resolution is poor unless data acquisition is underway.
Another potential problem is that there is a delay after scan initiation before data are acquired. This delay does not occur during advancement from one dynamic frame of data acquisition to another, however. Thus, it was necessary to modify the scan acquisition protocols to include two dynamic frames: Frame 1 had no preset duration and was begun 5–10 seconds prior to radiotracer administration. Frame 2 was preset to be 40 seconds in duration and was begun by the scanner operator when the coincidence events in the field of view sharply increased and indicated the arrival of the radiotracer bolus in the brain. Only the data from frame 2 was reconstructed and used in subsequent analysis.
The entire PET examination could be performed within 1 hour because of the short half-life (122.2 seconds) of 15O. All radionuclides were produced in the Washington University cyclotron facility (22,23).
Image Processing and Analysis
All images were reconstructed by using filtered backprojection and scatter correction with a ramp filter at the Nyquist frequency. They were then filtered with a three-dimensional Gaussian filter to a uniform resolution of 16 mm full width at half maximum. These images subsequently were transformed to stereotactic atlas space (24) by using the lateral skull radiograph and the transmission scan. This was done to allow reproducible placement of regions of interest.
When combined with the arterial time-activity curve data and the hematocrit values, the 15O-labeled carbon monoxide image provided the quantitative regional measurement of cerebral blood volume on a pixel-by-pixel basis (25). A regional map of quantitative cerebral blood flow was generated from the 15O-labeled water image by using arterial time-activity data (21,26,27). The 15O-labeled oxygen image provided the quantitative regional measurement of OEF, once combined with data from processed cerebral blood flow and cerebral blood volume images and the arterial time-activity curve information (17). Images depicting cerebral metabolic rate for oxygen were generated as the product of OEF, cerebral blood flow, and PaO2 content (17).
The count-based OEF image was generated as the ratio image of the counts in the filtered and atlas-transformed 15O-labeled oxygen and 15O-labeled water images, normalized to a whole-brain mean of 0.40—this value was the mean whole-brain quantitative OEF measured in the 68 patients with arterial time-activity curve data. This was performed in the following manner.
First, the normalization factor was calculated: (a) A template was created by using brain edges defined by isointensity contours at approximately 30% of the whole-brain maximum counts of the 15O-labeled water image to eliminate extracranial noise—the water image was used rather than the oxygen image because of the greater number of counts (ie, less noise); (b) this template was saved and used to mask the extracranial noise in the 15O-labeled oxygen image; (c) whole-section mean counts for 21 PET sections that included the middle cerebral arterial territories were calculated from both the 15O-labeled water and 15O-labeled oxygen scans by using the saved templates to mask the extracranial noise; (d) the ratio of the 15O-labeled oxygen counts to the 15O-labeled water counts was calculated; and (e) 0.40 was divided by this ratio to yield the normalization factor (0.40 = normalization factor x [whole-brain 15O-labeled oxygen counts/whole-brain 15O-labeled water counts]).
Next, a second mask was created in a similar manner to eliminate extracranial noise from the 15O-labeled oxygen images and 15O-labeled water images prior to the pixel-by-pixel division of the counts in the images: (a) A mask image was created from the 15O-labeled water image by using the same threshold percentage used for the template in the calculation of the normalization factor and in an identical manner (brain edges identified by means of isointensity contours at a specified threshold percentage of maximum whole-brain counts), and (b) this mask was applied to the 15O-labeled oxygen image.
Finally, the counts in the masked 15O-labeled oxygen image were multiplied by the normalization factor and divided by the counts in the masked 15O-labeled water image on a pixel-by-pixel basis. This resulted in the count-based OEF image used in the subsequent regional analyses (Fig 2).
For each patient and healthy volunteer, seven separate spherical regions of interest 19 mm in diameter were placed in the territory of the middle cerebral artery in each hemisphere. Each region included gray and white matter and was placed by using stereotactic coordinates (1,28). Areas of prior infarction were identified by two investigators (C.P.D., W.J.P.) by review of images depicting cerebral metabolic rate for oxygen (or 15O-labeled images for the 13 patients without arterial data) and CT or MR images. Neither the regions within these areas nor the corresponding contralateral regions were used for analysis. Left-to-right ratios of mean hemispheric values of both quantitative OEF images and count-based OEF images were calculated for patients and control subjects. For each patient, the ratio was considered abnormal if it decreased beyond the range observed in the normal sample.
Data Analysis
Several analyses were performed. First, the ability of the count-based OEF ratio to predict subsequent stroke was examined. The 81 patients were divided into two groups on the basis of comparison with the range of normal values: those with abnormal count-based OEF ratios and those with normal count-based OEF ratios. The primary analysis compared the two groups with respect to the length of time before reaching the end point of ipsilateral stroke by means of Kaplan-Meier survival curves and the log-rank statistic.
Second, the count-based technique was compared directly with the quantitative method in predicting ipsilateral stroke. In these analyses, the image data of the 68 patients with arterial time-activity curves was processed by using both count-based and quantitative methods. Quantitative and count-based OEF ratios were compared for each patient by means of a paired Student t test.
Finally, receiver operating characteristic (ROC) curves for the prediction of subsequent ischemic stroke were generated by varying the threshold ratio used for a positive test for both quantitative and count-based techniques. This was done by progressively increasing the threshold for an abnormal test result from a left-to-right OEF ratio of 1.00 ± 0.01 (0.99 to 1.01) to 1.00 ± 0.28 (0.72 to 1.28) by increments of 0.01. A value of P < .05 was used as the criterion of statistical significance. Statistical analyses were performed with SPSS 7.0 (SPSS, Chicago, Ill) for Windows.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Normal Range
The range for left-to-right hemispheric ratios for the quantitative OEF observed in the 18 healthy subjects was 0.914–1.084. The normal range of left-to-right hemispheric ratios for count-based OEF was 0.935–1.062 in these subjects.
Results of Count-based OEF for All 81 Patients
Fifty of the 81 patients with symptomatic carotid occlusion had abnormal results, with left-to-right hemispheric count-based ratios that fell outside of the range observed in the healthy subjects. In all 50, the hemisphere with the increased count-based OEF was ipsilateral to the occluded carotid artery. All 13 ipsilateral ischemic strokes occurred in the 50 patients with an increased count-based OEF (P = .002). The sensitivity of the count-based method for the prediction of subsequent ipsilateral stroke was 100% (95% CI = -23%) (29) with use of the range observed in the healthy subjects to define the normal range. However, 37 of the 68 patients who did not have an ipsilateral stroke had increased OEF (specificity of 46% [95% CI: 40.5%, 51.5%]). The Kaplan-Meier cumulative failure curve is shown in Figure 3. The ROC curve generated by progressively increasing the threshold at which the count-based OEF ratio was defined as abnormal is shown in Figure 4.
|
|
Nine ipsilateral strokes occurred in the 68 patients with complete quantitative data. By using the normal range of values observed in the healthy volunteers to define the threshold, 31 of the 68 patients had abnormal results with use of quantitative OEF ratios. Seven ipsilateral ischemic strokes occurred in this group of 31 patients, compared with two strokes in the 37 patients with quantitative OEF ratios within the normal range (P = .025; sensitivity of 78% [95% CI: 64.5%, 91.5%]; specificity of 59% [95% CI: 52.2%, 65.8%]). The count-based OEF ratio was less specific (specificity of 48% [95% CI: 41.2%, 54.8%]) and more sensitive (sensitivity of 100% [95% CI = -33%]) than the quantitative method. Forty-one patients had abnormal results, and all strokes occurred in this group (P = .005). Kaplan-Meier cumulative failure curves of the quantitative and count-based analyses for these 68 patients are shown in Figure 5a and 5b, respectively.
|
|
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Why does the count-based method presented in this article perform as well as, if not better than, the quantitative OEF technique in the prediction of subsequent stroke? The data from the count-based OEF image are linearly proportional to the quantitative OEF image except for the small contributions from intravascular oxygen and recirculating labeled water.
The equation for quantitative OEF developed by Mintun et al (17) is as follows:
t1t2 CartH20 (t) * e-ktdt is the activity due to the recirculating labeled water owing to 15O-labeled oxygen metabolism, where CBF is cerebral blood flow, t1 and t2 are the scanning time, and CartH20 is the time-dependent concentration of 15O-labeled water in the arterial blood. The term R x CBV x t1t2 CartO2 (t)dt defines the post midcapillary activity of the intravascular 15O-labeled oxygen where R is the ratio of small-vessel to large-vessel hematocrit values used in the calculation of CBV, CBV is the cerebral blood volume, and CartO2 is the time-dependent concentration of 15O-labeled oxygen in the arterial blood.
The two terms in the denominator of the equation are the blood flow delivering the oxygen (CBF x t1t2 CartO2 [t] * e-ktdt) minus the intravascular 15O-labeled oxygen (0.835 x R x CBV t1t2 CartO2 [t]dt). This method requires separate measurements of cerebral blood flow and cerebral blood volume to yield an accurate quantitative measurement of OEF. It has been validated in nonhuman primates by using intracarotid injections of 15O-labeled oxygen (17). In addition, it has been rigorously validated in nonhuman primates against direct measurements of arteriovenous oxygen difference in pathologic conditions (30).
The contribution of the recirculating labeled water of metabolism and the intravascular labeled oxygen to the counts measured in the brain is small in areas of normal cerebral oxygen metabolism (discussed later). Without these terms, the equation for OEF becomes the observed PET counts from the 15O-labeled oxygen image divided by the cerebral blood flow delivering the oxygen.
In the tissue autoradiographic technique of blood flow measurement used in our laboratory, the relationship between the tissue concentration of 15O-labeled water to the local blood flow is almost linear (21,26). In the count-based OEF method, therefore, we use the regional PET counts observed during the 15O-labeled oxygen scanning for the numerator and the regional PET counts observed during the 15O-labeled water scanning for the denominator: count-based OEF = PETobs (15O-labeled oxygen)/PETobs (15O-labeled water). Because we are comparing the ratio of the counts from the 15O-labeled oxygen images and the counts from the 15O-labeled water images from one hemisphere to the other, rather than using the absolute values for comparison between patients, the left-to-right ratio for the count-based estimate of OEF should be very similar to the left-to-right ratio of quantitative OEF.
In the first PET methods for OEF measurement, the ratio of counts from a steady-state or equilibrium 15O-labeled oxygen image to a steady-state 15O-labeled water image was used. This method was proposed by Jones and co-workers (31) before implementation and validation by Frackowiak et al (32). In this technique, blood flow is measured by using a continuous inhalation of 15O-labeled carbon dioxide. The labeled carbon dioxide is converted to labeled water by carbonic anhydrase in the lung and enters the blood. A steady-state situation in which the delivery of labeled water to the brain becomes equal to the washout and decay of tracer is reached after several half-lives.
The measured activity in the brain at this point is directly proportional to the blood flow. Quantitative values are provided by converting PET counts to milliliters per gram per minute by means of an arterial sample for the measurement of activity in the blood. A similar steady-state scan is obtained during the continuous inhalation of 15O-labeled oxygen. The ratio of the counts obtained during the oxygen scan to the counts measured during the water scan is linearly proportional to the OEF, except for the contribution of intravascular activity.
Baron and co-workers (18) compared the results of the steady-state 15O PET technique, uncorrected for intravascular 15O, with direct measurements of arteriovenous oxygen differences while changing PaCO2 in baboons. They found a linear relationship between the measured PET OEF and the arteriovenous oxygen difference. A consistent 13.3% overestimation of OEF by PET was found: OEF (arteriovenous oxygen difference) = 1.009 OEF (PET) + 13.3 (P < .001). To our knowledge, the first published description of misery perfusion used this uncorrected count-based technique of OEF measurement (2). In subsequent steady-state PET studies (19,33) that incorporated a correction for the intravascular 15O-labeled oxygen, the mean percentage of overestimation of actual quantitative OEF in uninfarcted tissue by using the uncorrected count-based method was 11%–14% in gray matter and 8%–9% in white matter.
The ROC analysis demonstrates similar sensitivity and specificity of the quantitative and count-based methods. In fact, the area under the curve is greater with the count-based technique. Thus, the count-based method performs as well as, if not better than, the quantitative technique in the prediction of subsequent ipsilateral stroke. With the relatively small number of patients examined with both methods with subsequent strokes at follow-up (nine patients), however, we cannot conclude with any confidence that the count-based OEF technique is better than the quantitative OEF method (34).
One aspect of the count-based OEF method that may provide greater accuracy is the use of the normalizing factor. Quantitative OEF values can vary widely between subjects with normal cerebral hemodynamics but are relatively uniform within individual patients. For example, one subject may have a quantitative OEF of 0.40 in both hemispheres, and another may have a value of 0.30. These absolute values of OEF are within the range of normal. Because of this normal variation in absolute values, we use the hemispheric ratio of OEF to identify abnormally increased OEF. An individual with a mean OEF of 0.30 in the left middle cerebral arterial territory and 0.40 in the right would have a very abnormal OEF ratio: left-to-right ratio of 0.30/0.40 = 0.75, well outside of the normal range.
However, in patients with low absolute values of quantitative OEF, the left-to-right ratio is more likely to fall beyond the normal range than in patients with higher OEF values. For example, a mean OEF of 0.40 on the left and 0.42 on the right would be categorized as normal (ratio = 0.95) while a mean OEF of 0.20 on the left and 0.22 on the right (ratio = 0.909) would be categorized as abnormal. In other words, although the absolute difference in quantitative OEF between the hemispheres is the same in both patients, the relative difference is greater in the patient with lower absolute values. By normalizing the count-based OEF image to a whole-brain mean of 0.40 before calculating the hemispheric ratio, this potential bias in categorization of patients with low absolute values is avoided.
If a multicenter trial is organized to test the effect of surgical revascularization in selected patients with high OEF, the count-based method presented here would hold several advantages over the quantitative technique. First, it may be possible to successfully enroll more patients. It may not be possible or safe to place an arterial catheter in many patients with carotid occlusion. This occurred in six of the 13 patients in whom quantitative processing could not be completed. Also, technical problems may render the arterial time-activity curve data useless, as occurred in five of the remaining seven patients.
Second, processing the count-based images is less complicated. The processing of the PET images, arterial activity curves, and other necessary laboratory data (hematocrit values, etc) required to yield quantitative metabolic data is both time and labor intensive (17,27). Third, the count-based technique does not require the synthesis and administration of 15O-labeled carbon monoxide for a cerebral blood volume scan. Finally, issues of standardization of PET methodology between institutions would be greatly simplified by using this count-based technique, as the software programs used for quantitative metabolic processing would need to be installed on different computer systems.
In summary, we have demonstrated that a count-based method of OEF measurement can be used to predict stroke in patients with symptomatic carotid occlusion. The sensitivity and specificity of this technique for the identification of patients at risk for stroke are comparable to those of the more complicated quantitative method. If a multicenter trial of surgical revascularization for patients with symptomatic carotid occlusion is organized, the count-based method of OEF measurement presented in this article should be used to select patients for surgery. The optimal threshold of the count-based OEF ratio used to select patients for surgery in such a trial will need to be determined through cost-effectiveness analysis.
|
Footnotes |
|---|
Author contributions: Guarantor of integrity of entire study, C.P.D.; study concepts and design, W.J.P., C.P.D.; definition of intellectual content, W.J.P., C.P.D.; literature research, C.P.D., W.J.P.; clinical studies, W.J.P., R.L.G., C.P.D., S.M.F., K.D.Y.; experimental studies, C.P.D., W.J.P., T.O.V.; data acquisition, C.P.D., S.M.F., K.D.Y., W.J.P., R.L.G., T.O.V.; data analysis, C.P.D., W.J.P., T.O.V., N.R.S.; statistical analysis, C.P.D., W.J.P.; manuscript preparation, C.P.D.; manuscript editing, W.J.P.; manuscript review, W.J.P., R.L.G., T.O.V.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. A. Krohn, J. M. Link, and R. P. Mason Molecular Imaging of Hypoxia J. Nucl. Med., June 1, 2008; 49(Suppl_2): 129S - 148S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kobayashi, T. Kudo, T. Tsujikawa, M. Isozaki, Y. Arai, Y. Fujibayashi, and H. Okazawa Shorter Examination Method for the Diagnosis of Misery Perfusion with Count-Based Oxygen Extraction Fraction Elevation in 15O-Gas PET J. Nucl. Med., February 1, 2008; 49(2): 242 - 246. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kobayashi, H. Okazawa, T. Tsuchida, K. Kawai, Y. Fujibayashi, and Y. Yonekura Diagnosis of Misery Perfusion Using Noninvasive 15O-Gas PET J. Nucl. Med., October 1, 2006; 47(10): 1581 - 1586. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P J Kirkpatrick and I Ng Cerebral revascularisation: where are we now? J. Neurol. Neurosurg. Psychiatry, April 1, 2005; 76(4): 463 - 465. [Full Text] [PDF]
|
|
|
|
 |
|
 C. P. Derdeyn, T. O. Videen, K. D. Yundt, S. M. Fritsch, D. A. Carpenter, R. L. Grubb, and W. J. Powers Variability of cerebral blood volume and oxygen extraction: stages of cerebral haemodynamic impairment revisited Brain, March 1, 2002; 125(3): 595 - 607. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. P. Derdeyn, T. O. Videen, R. L. Grubb Jr., and W. J. Powers Comparison of PET Oxygen Extraction Fraction Methods for the Prediction of Stroke Risk J. Nucl. Med., August 1, 2001; 42(8): 1195 - 1197. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. P. Derdeyn, A. Khosla, T. O. Videen, S. M. Fritsch, D. L. Carpenter, R. L. Grubb Jr, and W. J. Powers Severe Hemodynamic Impairment and Border Zone-Region Infarction Radiology, July 1, 2001; 220(1): 195 - 201. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
