Advertisement

Abstract

Background Cerebral autoregulation can be impaired after ischemic stroke, with potential adverse effects on cerebral blood flow during early rehabilitation.

Objective The objective of this study was to assess changes in cerebral blood flow velocity with orthostatic variation at 24 hours after stroke.

Design This investigation was an observational study comparing mean flow velocities (MFVs) at 30, 15, and 0 degrees of elevation of the head of the bed (HOB).

Methods Eight participants underwent bilateral middle cerebral artery (MCA) transcranial Doppler monitoring during orthostatic variation at 24 hours after ischemic stroke. Computed tomography angiography separated participants into recanalized (artery completely reopened) and incompletely recanalized groups. Friedman tests were used to determine MFVs at the various HOB angles. Mann-Whitney U tests were used to compare the change in MFV (from 30° to 0°) between groups and between hemispheres within groups.

Results For stroke-affected MCAs in the incompletely recanalized group, MFVs differed at the various HOB angles (30°: median MFV=51.5 cm/s, interquartile range [IQR]=33.0 to 103.8; 15°: median MFV=55.5 cm/s, IQR=34.0 to 117.5; 0°: median MFV=85.0 cm/s, IQR=58.8 to 127.0); there were no significant differences for other MCAs. For stroke-affected MCAs in the incompletely recanalized group, MFVs increased with a change in the HOB angle from 30 degrees to 0 degrees by a median of 26.0 cm/s (IQR=21.3 to 35.3); there were no significant changes in the recanalized group (−3.5 cm/s, IQR=−12.3 to 0.8). The changes in MFV with a change in the HOB angle from 30 degrees to 0 degrees differed between hemispheres in the incompletely recanalized group but not in the recanalized group.

Limitations Generalizability was limited by sample size.

Conclusions The incompletely recanalized group showed changes in MFVs at various HOB angles, suggesting that cerebral blood flow in this group may be sensitive to orthostatic variation, whereas the recanalized group maintained stable blood flow velocities.

Acute ischemic stroke adversely affects cerebral autoregulation, a homeostatic mechanism that maintains adequate cerebral blood flow despite variations in systemic blood pressure and cerebral perfusion pressure.16 Cerebral autoregulation ensures sufficient blood flow with changes in body position that affect blood pressure, such as lying, sitting, standing, or walking.7,8 Orthostatic positions closer to vertical present a greater challenge to this mechanism, particularly when it is impaired after stroke.5,9 Current physical therapy guidelines emphasize early rehabilitation after ischemic stroke, including sitting, standing, and walking, to facilitate optimal functional recovery.1012 The clinical implications of changes in blood flow velocities in response to such orthostatic variations in the acute phase after ischemic stroke remain unclear.

The mechanisms contributing to cerebral autoregulation are complex, and no direct measure of this mechanism is currently available. Investigations of the physiological effects of orthostatic variation have used measures of cerebral blood flow velocity to indirectly reflect the maintenance of cerebral autoregulation. Transcranial Doppler (TCD) ultrasound is a valid measure of blood flow velocity in the major cerebral arteries and is an accepted index of cerebral autoregulation.6,1316 Transcranial Doppler ultrasound has been used in several studies investigating autoregulation.3,8,14,1719 However, TCD studies of orthostatic blood flow changes in patients with stroke are limited.20,21

Ischemic stroke typically is caused by occlusion of a cerebral blood vessel22 (commonly the middle cerebral artery [MCA]). Obstructed blood flow, over time, results in the development of a necrotic infarct core surrounded by the ischemic penumbra, an area of neural tissue with potentially reversible damage. Salvage of the penumbra is the focus of treatment for acute stroke. The delivery of intravenous thrombolytic drugs to dissolve clots and recanalize (reopen) blood vessels is now standard practice in the medical management of acute ischemic stroke.14,16,23,24 Although thrombolytic therapy is considered the best practice for achieving arterial recanalization, its success depends on early delivery and many individual patient factors. Intravenous thrombolytic therapy has been reported to achieve arterial recanalization in 46% of appropriate patients; in comparison, 24% of patients with acute ischemic stroke show spontaneous recanalization without receiving active revascularization treatment.25 Patients whose stroke-affected arteries are recanalized often display rapid neurological improvement and often progress readily to rehabilitation and good recovery. To date, studies have not specifically investigated the physiological effects of orthostatic changes in patients showing early recanalization, and the extent of any potential impairment in cerebral autoregulation is not clear.

One report9 described the use of TCD ultrasound to measure the impact of orthostatic changes on blood flow velocities in patients with acute ischemic stroke. Cerebral blood flow velocity increased when the head of the bed (HOB) was lowered from 30 degrees to 0 degrees, indicating an improvement in cerebral blood flow with the bed flat. This increase in blood flow velocity was accompanied by clinically observed neurological improvement. However, patients were excluded from the study if they showed recanalization, and measurements from the nonaffected MCA were not obtained. Thus, it is unknown whether blood flow velocities respond differently to orthostatic changes in patients who show recanalization or whether there are differences between hemispheres.

The primary aim of this study was to describe changes in cerebral blood flow velocities in response to variations in orthostatic positions at 24 hours after ischemic stroke. Secondary aims were to determine any differences in blood flow velocities with changes in positions between arteries that had recanalized and those that had not and between stroke-affected and nonaffected hemispheres. We hypothesized that patients whose arteries had not recanalized would display greater changes in blood flow velocities with orthostatic variations than patients whose arteries had recanalized and that the stroke-affected hemisphere would similarly experience greater changes in blood flow velocities than the nonaffected hemisphere. Establishing the impact of orthostatic changes for patients after ischemic stroke may contribute to more effective care and rehabilitation for patients with acute stroke.

Method

Participants

Patients referred to a tertiary care hospital from April to August 2009 participated in this study; the sample size was determined by the number of eligible patients presenting within the 5-month data collection time frame. All patients had acute ischemic stroke with a duration of less than 6 hours since symptom onset and were considered for intravenous thrombolytic therapy.26 Patients enrolled in the study met the following criteria: 18 years of age or older, anterior circulation ischemia on computed tomography (CT) perfusion imaging, a National Institutes of Health Stroke Scale (NIHSS) score of 3 or greater, and a prestroke modified Rankin Scale score of 3 or less. The NIHSS is a nonlinear ordinal scale with scores from 0 to 42 that provides a valid and reliable assessment of acute stroke-related deficits.27,28 A score of 3 or greater indicates a stroke of sufficient severity to be considered for thrombolytic therapy. The modified Rankin Scale provides a measure of functional disability and physical dependence29; the score was obtained from family interview upon a patient's presentation to the hospital. A score of 3 or less indicates no significant preexisting disability.

Patients were excluded from participation if CT imaging revealed posterior circulation or hemorrhagic stroke, if temporal bone windows were insufficient to enable ultrasound penetration, or if they had substantial comorbidities that were life threatening or that limited their ability to lie flat in bed. In addition, patients were excluded if they displayed symptomatic orthopnea at the time of TCD data collection, for example, secondary to chronic obstructive pulmonary disease, asthma, or congestive cardiac failure. Patients unable to tolerate the TCD ultrasound procedure because of agitation were also excluded.

Written informed consent was obtained from all patients' families before recruitment into the study. Figure 1 shows the progression of participants through the study.

Figure 1.

Flow diagram showing the progression of participants through the study. CT=computed tomography, HOB=head of bed elevation, MCA=middle cerebral artery, MR=magnetic resonance, NIHSS=National Institutes of Health Stroke Scale, TCD=transcranial Doppler ultrasound. *Eligibility for thrombolytic therapy was described by Quain et al.26

Computed tomography or magnetic resonance angiography at 24 hours after stroke was used to determine arterial recanalization status and group participants accordingly. Complete recanalization was defined as full reopening of the affected artery and restoration of normal blood flow velocity (mean flow velocity [MFV] in a normal MCA is approximately 55 cm/s [SD=12]30,31). Incomplete recanalization was defined as partial or complete MCA occlusion and incomplete restoration of blood flow velocity.

Study Design and Protocol

This observational study followed a repeated-measures design. The dependent variable was MCA blood flow velocity, measured by TCD ultrasound. The manipulated independent variable was the angle of HOB elevation: 0, 15, and 30 degrees.

Equipment.

Blood flow velocities were measured by TCD ultrasound examination with a digital power-motion Doppler unit (PMD 100; Spencer Technologies, Seattle, Washington) and 2-MHz pulsed-wave diagnostic transducers at depths of 40 to 65 mm. A 10-mm area of insonation enabled optimal waveform visualization. Bilateral transducers were attached to the participant with a Marc 600 head frame (Spencer Technologies), positioned over the temporal bone windows, and angled to detect the optimal MCA signal. Blood flow velocities were measured at 5-mm increments along the MCA to detect the site of occlusion. The area of insonation then was positioned in the mid-M1 MCA segment in participants with complete recanalization and at the point just proximal to the site of occlusion where blood flow was detected in participants with incomplete recanalization. The head frame was fitted and all measurements were obtained by the same experienced neurosonographer to maintain consistency throughout the procedure.

The angle of HOB elevation was standardized with a Plurimeter (Australasian Medical & Therapeutic Instruments, Albany Creek, Queensland, Australia), a liquid pendulum inclinometer. The Plurimeter was attached to the bed frame with Velcro (Velcro USA Inc, Manchester, New Hampshire) and calibrated against the horizontal position by one researcher before testing for each participant.

Procedure.

A rapid-response stroke team (a neurologist and a specialist nurse) examined the participants and completed a thorough neurological assessment. Plain head CT was used to exclude hemorrhagic stroke and was followed by CT perfusion imaging and angiography to determine the extent of ischemia and the site of occlusion. Thrombolytic therapy was administered subsequent to specific criteria being met.26,32 A specialist nurse qualified in administering the NIHSS determined a baseline NIHSS score during the initial neurological examination and repeated the test after 24 hours. Computed tomography or magnetic resonance angiography at 24 hours after stroke was used to determine recanalization status and group participants accordingly. The TCD ultrasound protocol then was used to investigate MCA blood flow velocities at various HOB elevations.

The TCD ultrasound protocol was initiated at an HOB elevation angle of 30 degrees because the participants routinely rested in this position. The participant's hip joints were aligned with the hinge of the bed to ensure consistency of body position when the HOB was lowered and raised. An acclimatization period of 5 minutes with continuous TCD observation ensured the stability of blood flow velocities in the participants, who were expected to have impaired autoregulation because of stroke; autoregulation typically responds rapidly to positional changes in people who are healthy.1,21 Early rehabilitation includes prolonged periods of upright positions, such as sitting and standing; hence, monitoring the stabilization of cerebral blood flow in these positions was important. The acclimatization period was followed by 1 minute of data recording. Next, the HOB was lowered to 15 degrees; this step was followed by a 5-minute acclimatization period and then 1 minute of data recording. This process was repeated at an angle of HOB elevation of 0 degrees (bed flat). This protocol provided 1 minute of continuous data for analysis per HOB angle per participant.

The safety of the participants was ensured by beginning the protocol at the normal resting HOB elevation (30°) and progressively lowering the bed to the horizontal position. Blood flow velocity was expected to be greater in the horizontal position than at higher angles of HOB elevation in participants in the acute phase after ischemic stroke9; therefore, the horizontal position potentially posed less risk of ischemic damage than the participants' normal resting position.

Outcome Measures

Transcranial Doppler ultrasound produced a 1-minute recording of flow velocity waveforms. The MFV was determined for each waveform, and the values were then averaged over the 1-minute recording period after acclimatization at each HOB angle. The MCA MFV was the primary outcome measure used to compare differences between HOB angles. The change in MFV (from 30° to 0°) was used to compare differences between affected and nonaffected MCAs both within and between the recanalized and incompletely recanalized groups.

To ensure participant safety, a pulsatility index (a TCD measure of resistance to flow) was monitored throughout the procedure. A high pulsatility index reflects elevated resistance to flow and may be indicative of increased intracranial pressure.33 However, complications of increased intracranial pressure were not expected within the first 24 hours because increased intracranial pressure usually peaks at 48 hours after ischemic stroke or later.20,34,35 To further monitor participant safety, mean arterial pressure, heart rate, and blood oxygen saturation (Spacelabs Healthcare telemetry, model 90369; Spacelabs Healthcare, Issaquah, Washington) were concurrently measured during data collection to identify any physiological changes with manipulation of the HOB angle.

Data Analysis

Analysis of TCD waveforms was performed by reviewing the continuous TCD data files (Fig. 2). The files were assessed at a time and a location separate from those for participant data collection by an experienced neurosonographer and a neurologist, and the MFVs from each assessor were averaged to obtain a single value that was used in further analysis. Both assessors were unaware of participants' identification, diagnosis, and CT images to ensure unbiased and accurate interpretation of waveforms. The MFV was calculated for each waveform over the 1-minute recording period. Friedman tests were used to compare MFVs at the 3 HOB angles for affected and nonaffected MCAs in the incompletely recanalized and recanalized groups, and Wilcoxon signed rank tests were used to determine differences between pairs of HOB angles. Mann-Whitney U tests were used to compare changes in MFVs (from those at 30° to those at 0°) between groups and between affected and nonaffected sides within groups. Statistical analyses were performed with PASW Statistics 18.0 (SPSS Inc, Chicago, Illinois).

Figure 2.

Sample blood flow velocities in affected and nonaffected middle cerebral arteries (MCAs) at 30 and 0 degrees of elevation of the head of the bed (HOB) in a participant in the recanalized group and a participant in the incompletely recanalized group. Blood flow velocity in the affected MCA in the participant in the incompletely recanalized group was greater at 0 degrees than at 30 degrees, whereas in the nonaffected MCA, flow remained relatively unchanged.

Role of the Funding Source

No funding was required for this research, and there are no financial conflicts of interest.

Results

Of 35 patients who were screened for inclusion, 8 patients (5 women and 3 men; median age=66 years; interquartile range [IQR]=47.0 to 74.5 years) were enrolled in the study (Fig. 1). Five participants (4 women and 1 man) received thrombolytic therapy after being selected as eligible for this therapy according to current clinical guidelines.26 Angiography at 24 hours after stroke determined that 4 participants showed complete recanalization of the affected artery (recanalized group) and that 4 participants showed partial recanalization (incompletely recanalized group). No participants showed persistent complete occlusion at 24 hours. Participant demographics and NIHSS scores are shown in Table 1.

Table 1.

Participant Demographics at Baseline and 24-Hour Follow-upa

Blood Flow Velocities

The MFVs for individual participants are shown in Table 2. The stroke-affected MCA in the incompletely recanalized group was the only artery to demonstrate a significant difference in MFVs at various HOB angles (30°: median MFV=51.5 cm/s, IQR=33.0 to 103.8; 15°: median MFV=55.5 cm/s, IQR=34.0 to 117.5; 0°: median MFV=85.0 cm/s, IQR=58.8 to 127.0) (P=.039). There were no significant differences for nonaffected MCAs, for the stroke-affected MCA in the recanalized group, or between pairs of HOB angles for any group or hemisphere.

Table 2.

Average Valuesa for Stroke-Affected and Nonaffected Middle Cerebral Arteries by Angle of Elevation of the Head of the Bed (HOB) and Recanalization Status at 24 Hours After Stroke

Figure 3 shows the observed change in the MFV (from that at 30° of HOB elevation to that at 0° of HOB elevation) for the stroke-affected MCA in the incompletely recanalized group. For the stroke-affected MCA in the incompletely recanalized group, there was a median increase in the MFV (from that at 30° of HOB elevation to that at 0° of HOB elevation) of 26.0 cm/s (IQR=21.3 to 35.3), whereas in the recanalized group, there was no observable change (median difference=−3.5 cm/s, IQR=−12.3 to 0.8) (P value for between-group difference for MFV change=.021). The MFVs in the nonaffected MCAs in both groups also did not change significantly (median difference in MFV between 30° and 0° of HOB elevation in the incompletely recanalized group=−4.0 cm/s, IQR=−9.0 to 4.0; median difference in the recanalized group=−1.3, IQR=−3.9 to 1.8) (P value for between-group difference for median MFV change=.564). The difference between affected and nonaffected hemispheres in the change in MFV (from 30° of HOB elevation to 0°) was significant for the incompletely recanalized group (P=.020) but not for the recanalized group (P=.309).

Figure 3.

Box plots illustrating differences between affected and nonaffected middle cerebral arteries (MCAs) by group (recanalized or incompletely recanalized) in changes in mean blood flow velocities when the head of the bed was lowered from 30 degrees to 0 degrees at 24 hours after ischemic stroke.

Safety

No adverse events occurred during testing. There were no marked differences in the pulsatility index at the 3 HOB angles. Median mean arterial pressures (all participants) were 98.5 mm Hg (IQR=90 to 108.5), 97.5 mm Hg (IQR=90.5 to 105), and 95.5 mm Hg (IQR=87.5 to 103.5) at HOB angles of 30, 15, and 0 degrees, respectively. In addition, no clinically significant changes in heart rate or blood oxygen saturation were observed.

Discussion

The main finding of the present study was that MFVs in stroke-affected MCAs with incomplete recanalization increased as the HOB was lowered from 30 degrees to 0 degrees. No statistically significant changes were observed in nonaffected MCAs or MCAs with complete recanalization. The lack of a significant orthostatic variation in blood flow velocity after CT angiographic evidence of recanalization suggests intact autoregulation in the territory of the affected MCA. The lower blood flow velocity with HOB elevation in participants who did not show CT angiographic evidence of recanalization raises the possibility that, in this situation, the affected MCA territory may have had a decreased capacity to autoregulate and adapt to orthostatic “stress.” The implications for rehabilitation are that the early use of upright postures may place greater orthostatic stress on circulating blood volumes in participants who show evidence of incomplete recanalization. The findings suggest that implementing a routine CT examination at 24 hours after stroke may improve clinical decision making during early rehabilitation.

Interestingly, in our sample, 3 of the 4 participants who showed complete arterial recanalization did not receive thrombolytic therapy, whereas all participants in the incompletely recanalized group retained partial arterial occlusion despite the administration of thrombolytic therapy. Although these results are contrary to expectations for thrombolytic therapy, our small sample suggests that they may not be inconsistent with documented outcomes for this therapy in the literature.25 The lack of recanalization in participants receiving thrombolytic therapy in the present study probably reflects individual responses to treatment in our small sample and does not represent a limitation to the external validity of the present study.

Comparisons With Previous Research

Few studies have investigated the effects of changing orthostatic positions on cerebral blood flow velocities in acute ischemic stroke. Two reports9,21 documented the investigation of TCD ultrasound–measured MFVs with variations in the angle of the HOB of 0 to 30 degrees. Both studies identified a greater MFV in the stroke-affected MCA at 0 degrees of HOB elevation than at 30 degrees of HOB elevation, although important differences in patient selection and study design limit comparisons with the present study. One of these studies, by Wojner-Alexandrov et al,9 excluded patients who showed recanalization. This factor limits the generalizability of their findings, particularly with regard to future populations, because continued improvements in the efficacy of thrombolytic therapy are likely to increase rates of arterial recanalization. In the other study, Schwarz et al21 used bilateral TCD ultrasound measurements and identified a significant difference in blood flow velocities between affected and nonaffected hemispheres but no significant change in the nonaffected MCA. The findings for the incompletely recanalized group in the present study agree with the results of Schwarz et al,21 suggesting that cerebral perfusion may be impaired predominantly in the stroke-affected hemisphere. However, Schwarz et al21 did not categorize patients according to arterial recanalization, a factor that limits comparisons with the present study. Interestingly, in the present study, we identified no significant changes in MFVs with orthostatic variations in either hemisphere in participants who showed complete recanalization, suggesting greater stability of blood flow in these participants.

Physiological Rationale

In the present study, we identified large increases in MFVs in participants with incompletely recanalized arteries when the HOB was lowered to horizontal. This result suggests that the vasodilatory responses necessary to maintain constant cerebral blood flow in body positions approximating horizontal were impaired in these participants. The clinical impact of MFVs that are greater than normative values is uncertain. Increased MFVs beyond 48 hours after stroke onset are associated with hyperperfusion injury and hemorrhagic transformation, whereas hyperperfusion in the first 48 hours appears to be associated with neurological improvement without increased risk of hemorrhage.36 Furthermore, increased MFVs with the bed in the flat position were previously associated with clinically significant neurological improvement in 3 of 20 patients who did not show recanalization.9 Previous studies suggested that adopting a resting position of 0 degrees of HOB elevation may be beneficial for patients who do not show recanalization,9 as long as the intracranial pressure remains stable.21 In contrast, the effect on cerebral blood flow of raising the HOB above 30 degrees has not yet been studied. However, the association between body positions and MFVs in the acute phase after ischemic stroke in patients whose arteries remain occluded supports the biological rationale that orthostatic positions closer to vertical may present a potential risk of cerebral hypoperfusion. The clinical impact of large decreases in blood flow velocity in patients with acute stroke is unknown. One report37 documented that a decrease in blood flow velocity of greater than 70% during carotid clamping for endarterectomy was predictive of adverse clinical events, although other studies38,39 identified decrements of 85% occurring without clinical consequences.

Implications for Rehabilitation

Physical therapy guidelines currently advocate early rehabilitation after ischemic stroke to facilitate the optimal recovery of physical function.10,11 However, the most appropriate point at which to begin rehabilitation is not specified. Typically, patients in the acute stroke unit where the present study was conducted begin rehabilitation, including sitting, standing, and walking, 4 or 5 days after stroke. Other authors12 investigated the effects of very early rehabilitation, commencing within 24 hours after stroke. Phase II of the AVERT (Very Early Rehabilitation Trial for Stroke) study40 identified similar numbers of deaths and levels of disability at 3, 6, and 12 months of follow-up in groups receiving either very early rehabilitation (getting patients out of bed within the first 24 hours) or standard care. However, the impact of rehabilitation on blood flow velocities during this acute phase is uncertain.

In the present study, participants who showed recanalization early after stroke maintained greater stability of blood flow velocities during orthostatic changes at low angles of HOB elevation than participants who showed incomplete recanalization. Participants who showed recanalization may have retained blood flow stability in orthostatic positions closer to vertical. Conversely, in participants whose recanalization was incomplete, blood flow velocities were sensitive to orthostatic changes, even at low angles of HOB elevation. Therefore, these participants may have been more vulnerable to cerebral hypoperfusion in orthostatic positions closer to vertical. Follow-up angiography to determine arterial recanalization after stroke is not currently part of routine clinical practice. However, the apparent impact that arterial recanalization has on blood flow responses to orthostatic changes suggests a potential for imaging services to play a greater role in informing clinical decisions pertinent to rehabilitation in the future.

Limitations

The sample size in the present study was small and limited the power of statistical calculations. Furthermore, 1 participant in the incompletely recanalized group (participant 2, Tab. 2) displayed significantly greater MFVs in the stroke-affected hemisphere than did the other participants. The implications of increased MFVs are uncertain. Previous studies indicated the association of positive neurological outcomes with increased MFVs with the bed in the flat position.9 However, potentially detrimental effects of increased MFVs in patients with impaired cerebral autoregulation cannot be excluded. Thus, the inclusion of participant 2 in the present study means that the conclusions should be interpreted with caution.

Another limitation of the present study was the exclusion of patients who were unable to tolerate horizontal positioning because of symptomatic orthopnea. This factor may limit the generalizability of the findings, but it was necessary to ensure the safety of the participants. Additionally, TCD ultrasound may introduce some error because the measurements are operator dependent. However, TCD ultrasound is widely accepted as a safe and reliable noninvasive technique for measuring cerebral blood flow velocity. Consistency and accuracy in the present study were ensured by the use of one experienced neurosonographer.

Conclusion and Future Directions

In the present study, when the HOB was lowered to horizontal, blood flow velocities in affected MCAs that had not completely recanalized increased; there were no statistically significant changes in arteries that had completely recanalized. Participants who showed complete recanalization appeared to demonstrate greater stability of blood flow velocities, a result that may have been representative of a normalization of cerebral autoregulation. Participants who did not show complete recanalization, however, displayed significant changes in blood flow velocities with manipulation of the HOB between 0 and 30 degrees. Future studies with larger sample sizes are indicated to validate the findings of the present study as well as to examine the physiological effects and clinical implications of orthostatic positions closer to vertical in the acute phase after ischemic stroke.

Footnotes

  • All authors provided concept/idea/research design and consultation (including review of manuscript before submission). Ms Hunter and Dr Snodgrass provided writing. Ms Quain and Dr Parsons provided data collection. Ms Hunter, Dr Snodgrass, and Dr Parsons provided data analysis. Dr Parsons provided project management and institutional liaisons. Dr Parsons and Dr Levi provided participants and facilities/equipment.

  • Approval for this study was granted by the Hunter New England Area Health Service Human Research Ethics Committee (ref: 03/09/10/3.07).

  • An abstract presentation of this study was given at Stroke 2011: the 22nd Annual Scientific Meeting of the Stroke Society of Australasia; September 14, 2011; Adelaide, South Australia, Australia.

  • Received August 18, 2010.
  • Accepted June 16, 2011.

References

View Abstract