| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Experimental Studies |
1 From the Department of Radiology, University Hospitals, University of Leuven, Herestraat 49, B-3000 Leuven, Belgium (F.C., X.S., H.W., J.Y., G.M., Y.N.); Department of Radiology, Zhong Da Hospital, Southeast University, Nanjing, Jiangsu Province, China (F.C.); Department of Molecular and Vascular Biology, Faculty of Medicine, University of Leuven, Leuven, Belgium (Y.S., N.N.); Department of Pharmacology, Hamamatsu University, Hamamatsu, Japan (Y.S.); and the Affiliated Hospital of Weifang Medical University, Weifang, Shandong Province, China (X.S.). Received July 31, 2006; revision requested October 5; revision received October 12; final version accepted December 1. Address correspondence to Y.N. (e-mail: yicheng.ni{at}med.kuleuven.be).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEIMPLICATION FOR PATIENT CAREReferences |
|---|
Materials and Methods: The animal experiment complied with institutional regulations for laboratory animals. Stroke was induced in rats with photothrombotic occlusion of middle cerebral artery (MCA). T2-weighted, perfusion-weighted (PW), and diffusion-weighted (DW) MR imaging was performed 1 hour and 24 hours after occlusion. On the basis of PW and DW images at 1 hour, 49 rats with cortex and subcortex involvement and with perfusion-diffusion mismatch were randomly assigned into one of four groups: control group, group treated with 7.5 mg microplasmin, group treated with 10 mg/kg microplasmin, or group treated with 10 mg/kg tPA. Agents were intravenously injected 1.5 hours after occlusion. Infarct size and hemorrhagic transformation were assessed with MR imaging and histomorphologic findings. Neurologic deficit was scored. Measurements were statistically analyzed.
Results: There were 13 rats in the control group, 13 in the 7.5 mg/kg microplasmin group, nine in the 10 mg/kg microplasmin group, and 14 in the 10 mg/kg tPA group. Despite similar baseline perfusion-diffusion mismatch, histochemically defined total infarct volume was reduced from 25% ± 5 (standard deviation) in control group to 21% ± 2, 20% ± 4, and 20% ± 5 in 7.5 mg/kg microplasmin, 10 mg/kg microplasmin, and tPA groups, respectively, as similarly shown on T2-weighted, DW, and PW images at 24 hours (P < .05). Cerebral hemorrhage rate at 24 hours was higher in tPA group than in the other three groups. Bederson score of neurologic deficits was significantly reduced in treated groups compared with that in control group.
Conclusion: Perfusion-diffusion mismatch appeared useful in selecting candidates for thrombolytic therapy. Multiparametric MR imaging allowed noninvasive assessment of effects of microplasmin and tPA in rats; microplasmin had a significantly lower hemorrhagic rate.
© RSNA, 2007
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEIMPLICATION FOR PATIENT CAREReferences |
|---|
Microplasmin, a truncated form of plasmin, has been developed as a new therapeutic agent that combines a direct thrombolytic activity with a neuroprotective property (12,13). Results of previous studies have shown that microplasmin reduces brain tissue damage in stroke models of the mouse (14) and the rat (15). However, to our knowledge, direct comparison of tPA and microplasmin for their therapeutic effects in rats, the most commonly used species in cerebral ischemic studies, has not been reported.
In the treatment of acute stroke, thrombolytic agents target the brain tissue at risk (or the ischemic penumbra) that has been clinically defined as the ischemic, but still viable and salvageable, tissue if local perfusion is rapidly restored (16,17).
Multiparametric magnetic resonance (MR) imaging, including perfusion-weighted (PW) and diffusion-weighted (DW) imaging, has been widely applied for guiding and monitoring treatment in hyperacute stroke (2,8,9,11,16,18–20). The mismatch between relative lesion volume (rLV) on PW source images and that on apparent diffusion coefficient (ADC) maps (perfusion-diffusion mismatch) provides a practical and approximate measure of the tissue at risk and has been increasingly used for the evaluation of acute stroke. Therefore, the purpose of our study was to prospectively compare the therapeutic and hemorrhagic effects of microplasmin and tPA in stroke therapy by using multiparametric MR imaging in a photothrombotic stroke model of rats.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEIMPLICATION FOR PATIENT CAREReferences |
|---|
MR Imaging
MR imaging was performed with a 1.5-T unit (Sonata; Siemens, Erlangen, Germany) by using a commercially available 4-channel phased-array wrist coil. Rats were placed supinely into a plastic holder and were anesthetized with 2% isoflurane (Forene; Abbott, Louvain-la-Neuve, Belgium) by using the same breathing anesthesia system used at surgery. For each imaging sequence, three pilot images (sagittal, coronal, and transverse) were obtained first; subsequently, 12 coronal images were acquired with a section thickness of 2 mm and intersection gap of 0.2 mm.
Turbo spin-echo T2-weighted MR imaging was performed with a repetition time msec/echo time msec of 5860/100, field of view of 55 x 31 mm, and acquisition matrix size of 256 x 146. DW MR imaging (3000/80) was performed by using a two-dimensional spin-echo echo-planar sequence with a field of view of 140 x 70 mm and an acquisition matrix size of 192 x 96, resulting in an in-plane resolution of 0.7 x 0.7 mm. A parallel imaging technique—namely, a generalized autocalibrating partially parallel acquisition (GRAPPA; Siemens) with an acceleration factor of two was applied. ADC maps were calculated on the basis of three b values (0, 500, and 1000 sec/mm2). Dynamic susceptibility contrast material–enhanced PW MR imaging was performed by using a T2*-weighted (2000/40) echo-planar sequence in combination with the GRAPPA technique (Siemens) to increase spatial homogeneity. A dynamic image series of 100 measurements (12 images per measurement and 1200 images in total) resulted in an imaging time of 3 minutes 34 seconds, with a field of view of 156 x 43.5 mm and an image acquisition matrix size of 128 x 64, which led to an in-plane resolution of 0.7 x 0.7 mm. During the dynamic series, bolus injection of contrast material (gadodiamide, Omniscan; Amersham, Oslo, Norway) at a dose of 0.3 mmol per kilogram of body weight was started after the 30th imaging acquisition to ensure sufficient precontrast baseline images. The entire MR imaging protocol typically lasted 20 minutes for each animal at each time point. All MR imaging sequences were performed 1 hour and 24 hours after MCA occlusion.
Study Design and Procedure
This animal experiment had a randomized, blinded, and placebo-controlled study design. Animals with cortex involvement on PW source images at the central coronal section or the sixth section from the tentorium of the cerebellum and with perfusion-diffusion mismatch were entered into the study with the consensus of three observers (F.C., N.N., and Y.N., with 15, 4, and 16 years of experience, respectively) on the basis of visual evaluation, as previously described and validated in detail (15,19). The qualified rats were randomly assigned to one of the following treatment groups with blinded labels: a control group treated with solvent, a group treated with 7.5 mg/kg microplasmin, a group treated with 10 mg/kg microplasmin, and a group treated with 10 mg/kg tPA.
Both microplasmin and tPA were dissolved in saline with 33 mmol/L mannitol and 5 mmol/L citric acid (pH = 3.1). The control agent was prepared with the same solvents adjusted at equal quantity and pH (14,23). With blinded labels, 0.2 mL of each group's agent was intravenously administered in rats 1.5 hours after MCA occlusion by using a single bolus. No heparin was given.
Before sacrifice, neurologic deficits of the rats were assessed (Y.S., with 10 years of experience) by using the ranking method of Bederson et al (24). After we tested wrist and elbow flexion and postural reflex and circling for determining paw dysfunction, hemiparesis, and obtundation, each score was summed and represented as a single neurologic score (0–10).
Histochemical Staining and Histologic Examination
After in vivo studies, animals were euthanized with overdose of pentobarbital (Nembutal; Sanofi Sante Animale, Brussels, Belgium). Brains were excised and sliced into six coronal slices of 2 mm thickness from the anterior 3.5–13.5 mm by using a brain matrix (Agar Scientific, Stansted, England). After hemorrhage was registered, slices were incubated at 37°C in a 2% solution of 2,3,5-triphenyltetrazolium chloride (TTC) for 15 minutes; normal brain tissue stained brick red, and cerebral infarct remained pale. The brain slices were subsequently fixed with 10% formalin and processed with hematoxylin-eosin staining for microscopy.
The cortical and subcortical infarct volumes at TTC staining were obtained with image analysis software (Image J, version 1.34s; National Institutes of Health, Bethesda, Md). With cross-referencing TTC staining, microscopic cellular changes, such as neuron necrosis, neuropil vacuolation, and inflammatory infiltration, were observed qualitatively by consensus (J.Y., with 16 years of experience; F.C.; Y.N.).
Image Analysis
Image analysis was performed off-line on a Linux workstation by using dedicated software (Biomap; Novartis, Basel, Switzerland). For quantification of rLV on T2-weighted, DW, and PW images, the areas of the lesion and the entire brain were delineated by three authors in consensus (F.C., Y.N., Y.S.) with an operator-defined region of interest on all lesion-containing sections. For the delineation of ischemic lesions, ADC maps and PW source images were used because of the clearer lesion demarcation derived with the clinic MR imager. The lesion was contoured as a whole because it was difficult to separate the cortex and subcortex on rat brain images. The software automatically generated the lesion volume (LV) and brain volume (BV) for each animal, from which rLV was calculated as rLV = (LV/BV)·100%.
The mismatch between rLV values on PW source images and rLV values on ADC maps at 1 hour after MCA occlusion was defined as [(sPW – ADC)/sPW] > 10%, where sPW is PW source image, and was applied to justify animal inclusion on the basis of visual evaluation of PW and DW images before treatment.
The signal intensity changes on DW images and ADC maps were observed, and the relative ADC value of the total ischemic lesion (lesion ADC value divided by contralateral ADC value) on three representative sections was quantified at 24 hours to reflect the degree of cellular damage.
Software built into the MR unit (Siemens) was used to generate PW-derived parametric maps. A 25-mm2 region of interest containing 49 pixels of 0.72 x 0.72 mm was placed on the region contralateral to the region that contained the lesion, which included the origin of the right MCA to allow measurement of the arterial input function, from which the pixels representing the right MCA branch were selected (20) with consensus of three authors (F.C., Y.N., Y.S.).
Relative cerebral blood volume (rCBV) and relative cerebral blood flow (rCBF) values were derived as described by the following steps: (a) one central lesion-containing section was chosen from PW source images, which provided a clearer image than PW-derived maps and enabled separate delineation of the cortex and subcortex; (b) the cortical and subcortical regions of interest thus obtained from images of the lesional and contralateral brain were transferred to rCBF and rCBV maps; (c) readouts of rCBF and rCBV were automatically generated by the software; and (d) the percentage of lesional rCBF and rCBV in comparison to those in the contralateral normal sides was calculated (20).
To assess the extravasation of contrast agent at MR imaging as an in vivo indicator of disintegration of the blood-brain barrier, the spatial-temporal changes in T2* relaxation rate (ie, 
R2*) on dynamic susceptibility contrast-enhanced PW images in the region of cerebral ischemia were calculated as follows: R2*(t) = 1/T2*(t) = –ln[SI(t)/SI0]/TE, where SI(t) and SI0 are signal intensities at time t and before injection of contrast agent, respectively (5), and TE is echo time in milliseconds.
The multifocal cerebral hemorrhages on macroscopic brain slices were visually determined by consensus (N.N., Y.N., Y.S.). The rate of hemorrhage (as a percentage) was calculated as the number of hemorrhagic cases divided by the total number of cases in each group. The volume of hemorrhage was not measured because of technical constraints at both MR imaging and histomorphologic analysis.
Statistical Analysis
Parametric data were reported as means ± standard deviations, and nonparametric data were expressed as medians and ranges. Statistical analysis was performed with software (SPSS for Windows, release 13.0; SPSS, Chicago, Ill). A general linear model of univariate test was applied for global comparisons between groups and techniques, followed by least significant difference tests for multiple comparisons of lesion size and ratio. For comparisons of MR imaging measurements between 1 hour and 24 hours, repeat analysis of variance was performed. The 
2 test was performed to compare rates of hemorrhage occurrence, and the Mann-Whitney U test was performed to compare behavior scores between groups. A P value of less than .05 was considered to indicate a significant difference. Adequate statistical power was ensured for our study by increasing the number of cases in each group in comparison with that in one of our previous studies, which showed a calculated statistical power of 0.90 (15).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEIMPLICATION FOR PATIENT CAREReferences |
|---|
Neurologic Symptom Scores
At 24 hours, all animals in the control group exhibited neurologic impairment, with a median Bederson score of 7 (range, 4–9). For the 7.5 mg/kg microplasmin, 10-mg/kg microplasmin, and 10 mg/kg tPA groups, there was a significant reduction in median score—to 5 (range, 3–8), 4 (range, 3–6), and 4 (range, 3–7), respectively (P < .05).
MR Imaging Findings and Perfusion-Diffusion Mismatch at 1 Hour
Lesions on T2-weighted images 1 hour after occlusion were not measured because of the difficulty of delineation. Before treatment, rLV values defined at 1 hour on ADC maps and PW source images, respectively, in the control group (21% ± 4 and 26% ± 5) were not significantly different from those in the other groups, with rLV values in the 7.5 mg/kg microplasmin group of 22% ± 7 and 25% ± 7; in the 10 mg/kg microplasmin group of 19% ± 5 and 24% ± 4; and in the 10 mg/kg tPA group of 22% ± 4 and 26% ± 6 (P > .05).
The value of the perfusion-diffusion mismatch at 1 hour in rats of the control group (23% ± 17) was also similar to that in rats of both microplasmin-treated groups (29% ± 14 and 30% ± 14 in 7.5 mg/kg and 10 mg/kg microplasmin groups, respectively) and the tPA-treated group (22% ± 17) (P > .05) (Fig 1).
|
|
|
|
TTC Staining Findings at 24 Hours
rLV values measured at TTC staining at 24 hours in the cortex and the total lesion, respectively, were significantly reduced in the 7.5 mg/kg microplasmin group (8% ± 3 and 21% ± 2), 10 mg/kg microplasmin group (7% ± 4 and 20% ± 4), and 10 mg/kg tPA group (6% ± 3 and 20% ± 5) (P < .05) compared with those in the control group (12% ± 5 and 25% ± 5). For rLV values from the subcortical lesion, there was no significant change between the control group (13% ± 3) and the three treated groups (12% ± 3 for the 7.5 mg/kg microplasmin group, 11% ± 1 for the 10 mg/kg microplasmin group, and 12% ± 3 for the 10 mg/kg tPA group) (P > .05) (Fig 5).
|
Comparison of rCBV and rCBF between 1 Hour and 24 Hours
The rCBV and rCBF of the ischemic lesion in the cortex at 24 hours were significantly increased compared with those at 1 hour for each group (P < .05). For the lesion in the subcortex, there were no significant changes in rCBV and rCBF at 24 hours compared with those at 1 hour for each group (P > .05), except for a significant decrease in rCBF in the 7.5 mg/kg microplasmin group (P < .05). No significant changes for rCBV and rCBF between 1 hour and 24 hours at the cortex and subcortex were found between the three treated groups and the control group (Fig 6).
|
|
|
R2* at 24 hours was significantly higher in the 10 mg/kg tPA group (0.0014 sec–1 ± 0.0011) than in the 7.5 mg/kg microplasmin group (0.0005 sec–1 ± 0.0004), the 10-mg/kg microplasmin group (0.0002 sec–1 ± 0.0003), and the control group (0.0004 sec–1 ± 0.0004) (P < .001). As indicated in Figure 7, the R2* curve from the tPA group showed the largest tail compared with that in the three other groups (Fig 7).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEIMPLICATION FOR PATIENT CAREReferences |
|---|
R2*. In our study, we used the visual perfusion-diffusion mismatch as a tool to select appropriate animals for treatment. The calculation during imaging analysis showed that perfusion-diffusion mismatch occurred homogeneously among the four groups before treatment. The animals thus selected responded well to the therapy in the microplasmin and tPA groups. As reported for clinical studies (26), the use of perfusion-diffusion mismatch proved to be both feasible and helpful in this study in identifying potential responders to thrombolytic therapies.
Our study results also suggested that MR imaging findings could be used as a noninvasive end point to reflect the pathologic nature of the ischemic lesion. Although T2-weighted imaging alone was not sensitive at 1 hour, multiparametric MR imaging, including T2-weighted imaging, DW imaging, and PW imaging, at 24 hours accurately depicted infarction volume as proved at TTC staining for all groups of rats in this study, which is in line with results elsewhere (19,20). MR imaging outcomes in the control group support the findings of a previous invasive study that showed that ischemic infarction was fully developed at 24 hours in the same stroke model (27).
On ADC maps and PW source images, the expansion of the ischemic lesion between 1 hour and 24 hours was significantly lessened by using microplasmin and tPA, in contrast to findings in the control group. In animal research, it is important to determine whether the obtained difference is the result of different treatments or because of individual variations. In this study, despite the insignificant variation of initial lesion size between the control and treated groups, we documented the lesion by using MR imaging at 1 hour and 24 hours and subsequently compared the lesion expansion during this period; hence we used a more reliable method that minimized influence from individual variations. Therefore, the observed reduction in lesion expansion can be attributed to the vessel recanalization caused by thrombolytic agents (25,28).
With regard to the PW-derived functional parameters, cortical rCBV and rCBF between 1 hour and 24 hours were significantly increased in each group; this increase was not significantly different between the control and treated groups. In control rats, this was likely because of spontaneous reperfusion and/or collateral circulation 3–8 hours after MCA occlusion, as observed elsewhere with the same stroke model (29–32). However, such late reperfusion due to endogenous thrombolysis may not be efficient enough to rescue the reversibly ischemic tissue from infarction (27). On the contrary, reperfusion induced by using microplasmin and tPA likely occurred much earlier in this study because of the more prompt thrombolytic intervention at 1.5 hours after occlusion and hence more efficient restoration of blood flow. Reperfusion and restored tissue viability were almost always found in the cortex.
However, subcortical rCBV and rCBF between 1 hour and 24 hours decreased in each group, but this decrease was not significantly different between the control and treated groups. The observed poorer postischemic perfusion may be because the subcortical region is more vulnerable to ischemic injury (15) than the cortical region. Our MR imaging results showed that the ADC reversion observed at 24 hours always occurred in the subcortical lesion of all groups, which is consistent with the overall lower perfusion in this region. No significant difference in lesional ADC was found between the control and treated groups. At microscopy, the area with ADC reversion or increased water diffusion reflected colliquative necrosis with loss of cellular membranes as the diffusive barriers. All the above data suggest that an ischemic lesion in the subcortex is less affected by thrombolytic agents than one in the cortex.
It has been suggested that introduction of tPA aggravates degradation of microvascular barriers (ie, mainly basal lamina and blood-brain barrier) in ischemic injury (5). The breakdown of basal lamina—caused mainly by matrix metalloproteinase-9 induction by lipoprotein-receptor–relating protein (7,25,33,34)—is associated with hemorrhagic transformation (5). Consequently, exogenous contrast agents may leak from vessels and induce parametric alterations on MR images that are then exploitable as a tool for diagnosis of cerebral bleeding and blood-brain barrier disintegration. A high dose of intravenously administered paramagnetic gadopentetate dimeglumine induced T2* (1/R2*) shortening or the enhanced transverse relaxation rate R2* on dynamic susceptibility contrast-enhanced PW images. Therefore, abnormal R2* changes (R2*) reflect extravasation of gadopentetate dimeglumine or increased permeability of the blood-brain barrier (5). In our study, the R2* curve from the tPA group had the largest tail, which implied considerable leakage and contrast material retention in the brain tissue after the first pass (35), which in turn led to a significantly higher R2* than that in the microplasmin and control groups at 24 hours. Our results suggest that tPA treatment aggravated the blood-brain barrier damage that may herald a higher risk of hemorrhage.
The exact dose of thrombolytic drugs seems critical not only in regard to the antistroke effects but also in regard to the hemorrhagic side effect, and the optimal dose of each drug also varies among species of animals (15,23,25,36). For our study, we used 10 mg/kg tPA, which is approximately equivalent to its standard clinical dose of 1 mg/kg for patients. The doses of 7.5 and 10 mg/kg microplasmin were considered appropriate for depleting circulating 
2-antiplasmin to exert therapeutic effects without resulting in much safety concern. Indeed, all rats in the three treated groups had comparable antistroke effects. However, administration of 10 mg/kg tPA caused significant bleeding, as evidenced with histomorphologic findings, which were in agreement with findings of previous studies (5–8,23,36). Interestingly, there was a dose-dependent neural and/or vascular protective effect with microplasmin because a significantly lower hemorrhagic rate was found, particularly with a higher dose (10 mg/kg) of microplasmin. Such a potential benefit deserves further preclinical and clinical exploitations.
Although the detailed mechanisms are still unknown, the therapeutic effects of microplasmin may be realized on two levels. The first is the partial recanalization of the occluded vessels, as evidenced by the improved MR imaging parameters through direct thrombolysis (15) due to the neutralization of plasma 2-antiplasmin. The second is the nonthrombolytic neuroprotective effect, which is supported by the fact that administration of microplasmin decreased the infarct volume even in a mouse model with permanent MCA occlusion (14,25). Therefore, the improvement in neurologic deficits in our study may be attributed to both of these therapeutic effects of microplasmin.
Certain limitations existed in our study. First, because of constraints of the clinical MR imager, the field of view as an imaging acquisition parameter was not identical between MR imaging sequences, which hampered direct coregistration for correcting geometric distortion. Therefore, point-to-point matching between MR images was not feasible. As an alternative, relative values were applied to facilitate the comparison between animals and imaging methods. Second, volumetric quantification of intracerebral hemorrhage was not attempted. Instead, the rate of hemorrhage was compared between control and treated groups. Third, because of the limited accessibility of the clinical imager, we were unable to monitor the therapeutic effects at more time points between 1 hour and 24 hours.
In conclusion, perfusion-diffusion mismatch is a useful tool in selecting appropriate animals for stroke therapy. Both microplasmin and tPA shared a similar effect in preventing infarct lesion expansion with improved behavioral scores. However, the intracerebral hemorrhage rate was significantly lower in rats treated with microplasmin than in rats treated with tPA. A multiparametric MR imaging approach allows noninvasive assessment of the therapeutic effects and hemorrhagic transformation after administration of antistroke agents.
Practical application: Microplasmin may serve as a potential alternative to tPA in the treatment of patients with ischemic stroke, with comparable effects and lower hemorrhage rate.
|
ADVANCES IN KNOWLEDGE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEIMPLICATION FOR PATIENT CAREReferences |
|---|
|
IMPLICATION FOR PATIENT CARE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEIMPLICATION FOR PATIENT CAREReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: ADC = apparent diffusion coefficient • DW = diffusion weighted • MCA = middle cerebral artery • PW = perfusion weighted • rCBF = relative cerebral blood flow • rCBV = relative cerebral blood volume • rLV = relative lesion volume • tPA = tissue plasminogen activator • TTC = triphenyltetrazolium chloride
Authors stated no financial relationship to disclose.
See also Science to Practice in this issue.
Author contributions: Guarantors of integrity of entire study, F.C., N.N., G.M., Y.N.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, F.C., Y.S., N.N., Y.N.; experimental studies, F.C., Y.S., N.N., X.S., H.W., J.Y., Y.N.; statistical analysis, F.C., Y.S., N.N., H.W., Y.N.; and manuscript editing, F.C., Y.N.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEIMPLICATION FOR PATIENT CAREReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
