Fax +41 61 306 12 34E-Mail karger@karger.chwww.karger.com

Original Paper

Eur Neurol 2012;67:74–80 DOI: 10.1159/000333778

Which Cortical Area Is Related to the Development of Dysphagia after Stroke? A Single Photon Emission Computed Tomography Study Using Novel Analytic Methods

Ryo Momosaki Masahiro Abo Wataru Kakuda Go Uruma

Department of Rehabilitation Medicine, Jikei University School of Medicine, Tokyo , Japan

significantly lower in the dysphagia group. The highest area under the curve was found in Brodmann area 4. In this area, 80% sensitivity and 60% specificity for discriminating dys-phagia were achieved with an optimal cutoff value. Conclu-

sions: When analyzed with novel methods, SPECT imaging can be useful for predicting the risk of dysphagia and subse-quent aspiration in poststroke patients.

Copyright © 2011 S. Karger AG, Basel

Introduction

Dysphagia is one of the most frequently seen impair-ments in poststroke patients. The incidence of dysphagia has been reported to be 55% in acute stroke patients [1] . The presence of this impairment is significantly associ-ated with a decreased quality-of-life level [2] . Previously, brainstem lesions were considered to be a principle injury causing dysphagia [3] . However, an association between cerebral cortical lesions and the development of dyspha-gia has recently been suggested. Hamdy et al. [4, 5] and Martin et al. [6] reported that some cortical areas such as the primary somatosensory motor cortex, supplementa-

Key Words

Single photon emission computed tomography � Dysphagia � Stroke

Abstract

Purpose: The purpose of this study is to clarify cortical areas related to the development of dysphagia in poststroke pa-tients using novel analytic methods for single photon emis-sion computed tomography (SPECT) imaging. Subjects and

Methods: Twenty poststroke patients (age: 66.1 8 5.1 years) with a left hemispheric lesion were studied. According to clinical evaluation, patients were divided into a dysphagia group (n = 10) and a control group (n = 10). In each patient, measurement of regional cerebral blood flow (rCBF) was performed by SPECT imaging with a 99mTc-ethylcysteinate dimer. For the analysis, an easy Z-score imaging system and voxel-based stereotactic extraction estimation were ap-plied, with placing regions of interest segmented into the Brodmann area level. We compared rCBF in each area be-tween the two groups, and receiver operating characteristic analysis to calculate the area under the curve was also per-formed. Results: The rCBF in Brodmann areas 4 and 24 was

Received: June 16, 2011 Accepted: September 25, 2011 Published online: December 22, 2011

M. Abo, MD, PhD Department of Rehabilitation Medicine, Jikei University School of Medicine 3-25-8, Nishi-Shimbashi, Minato-Ku Tokyo 105-8461 (Japan) Tel. +81 3 3433 1111, E-Mail abo   @   jikei.ac.jp

© 2011 S. Karger AG, Basel0014–3022/12/0672–0074$38.00/0

Accessible online at:www.karger.com/ene

Dow

nloa

ded

by:

Uni

vers

ity o

f Tok

yo

13

3.11

.204

.54

- 11

/15/

2014

9:2

3:04

AM

Cortical Areas Causing Dysphagia after Stroke

Eur Neurol 2012;67:74–80 75

ry motor cortex, and anterior cingulated gyrus play an important role in swallowing function. Although they showed this relationship based on the findings of neuro-imaging, it has been pointed out that the regional cerebral blood flow (rCBF) values on neuroimagings such as sin-gle photon emission computed tomography (SPECT) and positron emission tomography can vary among research-ers due to differences in the skill of manually setting re-gions of interest (ROIs) [7] .

To solve these problems, two new software packages which allow stereotactic analysis of SPECT using voxel-based analysis [8] were recently developed and are now available in Japan. One is the easy Z-score imaging sys-tem (eZIS) employing statistical image analysis. Using eZIS, changes in rCBF can be shown as Z-values based on comparison with a normal image database [9] . Another is the voxel-based stereotactic extraction estimation (vbSEE) [10] . Using this software, rCBF data on eZIS can be automatically converted to Talairach coordinate spac-es. These software packages have been reported to auto-matically set ROIs on the neuroimaging more accurately and with more reliability than when the manual setting was applied. So far, however, no clinical study that inves-tigated which cortical lesion is associated with the devel-opment of dysphagia using these novel analytical meth-ods. It is expected that SPECT evaluation using these an-alytic software packages can clarify this clinical issue with higher reliability.

Therefore, the purpose of this study is to identify ce-rebral cortical regions related to the development of dys-phagia with the SPECT imaging analyzed with these nov-el statistical software packages. For the identified respon-sible legions, we also clarified the cutoff values of rCBF for diagnosing the presence of dysphagia.

Subjects and Methods

Subjects The local ethics review committee at Jikei University School

of Medicine approved the protocol of this case-control study, and written consent was obtained from each subject for the study. Twenty Japanese poststroke patients with left hemispheric lesions were consecutively recruited for this study and studied. As there were a lot of patients with a left hemisphere stroke, we only as-sessed patients with left hemispheric lesions for this study. All patients were right-handed and were admitted to our department for rehabilitation during the period between May 1, 2009, and May 31, 2010. The period between onset of stroke and the admis-sion was more than a week in all patients. On admission, the clin-ical diagnosis of stroke and the location of the lesion were con-firmed with brain CT or T2-weighted images of MRI. Stroke le-sions were identified by a radiologist. Patients with a history of

previous stroke, multiple cerebral lesions or diffuse white matter lesions on brain imaging, or apparent dysphagia prior to the ad-mission were excluded from the study. We defined diffuse white matter by the Age-Related White Matter Changes Rating Scale for MRI and CT [11] as follows: white matter lesions’ diffuse involve-ment of the entire region. Forty-six left hemisphere stroke pa-tients admitted to the rehabilitation department during this pe-riod and twenty-six patients were excluded. There were no pa-tients with aphasia and apraxia in either group.

Basically, this study was performed by comparing the rCBF in some cortical areas evaluated by SPECT between patients with dysphagia (dysphagia group) and without the symptom (control group). For the diagnosis of dysphagia, the modified water swal-lowing test (MWST), and the flexible endoscopic evaluation of swallowing (FEES) were applied within 7 days prior to SPECT imaging. The modified water swallowing test is a screening test for dysphagia [12] . Cold water (3 ml) was placed on the floor of the mouth using a syringe. If the patient was unable to swallow, a score of 1 was given. If the patient was able to swallow but experi-enced difficulty breathing after swallowing, a score of 2 was given. If the patient was able to swallow and experienced coughing or wet-hoarseness after swallowing, a score of 3 was given. Other-wise, the patient was asked to perform two dry swallows. If the patient was able to swallow the water but unable to perform either of the two dry swallows, a score of 4 was recorded. If the patient was able to complete the water and both dry swallows, a score of 5 was recorded. The entire procedure was repeated twice more and the final score was defined as the lowest score on any trial. FEES is an evaluation of the pharyngeal stages of swallowing with a flexible fiberoptic laryngoscope [13] . It compares well with the video fluoroscopic examinations [14] and was considered the gold standard. The patient was considered to have a dysphagic symp-tom and was assigned to the dysphagia group of the study when the MWST score was less than 4 and the result of FEES deemed food modification necessary. If the MWST score was over 4 or the FEES result did not call for food modification, the patient was as-signed to the control group of the study.

The swallowing function was evaluated employing the func-tional oral intake scale (FOIS) in which the swallowing function is evaluated based on ingested food forms and rated as levels 1–7. The swallowing function becomes favorable as the level rises. FOIS was assessed within 2 days before SPECT [15] . For the eval-uation of the general severity of stroke, the NIH stroke scale was assessed at the time of SPECT imaging.

Single Photon Emission Computed Tomography Each subject underwent SPECT evaluation for the measure-

ment of rCBF. As a SPECT system, a three-detector gamma cam-era (PRISM-IRIX; Shimadzu Medical Co., Kyoto, Japan) with a low-energy high-resolution collimator was used. The evaluation was performed 20 min after intravenous injection of 600 MBq of 99mTc-ethylcysteinate dimer, while lying down in the supine po-sition with eyes closed in a quiet room. The SPECT images were prepared with attenuation correction according to Chang’s meth-od [16] . The parameters for SPECT imaging were as follows: the step and shoot method, 50 s per angle, 72 angles, matrix size 128 ! 128, location window 140 keV, image voxel size 3.2 mm. For prefilter and absorption correction, the Ramp filter (order 8.0, cutoff 0.27) was used.

Dow

nloa

ded

by:

Uni

vers

ity o

f Tok

yo

13

3.11

.204

.54

- 11

/15/

2014

9:2

3:04

AM

Momosaki/Abo/Kakuda/Uruma

Eur Neurol 2012;67:74–8076

eZIS Application The eZIS program, which was developed by Matsuda et al. [9] ,

was used to investigate the relative rCBF changes in all subjects. Using the SPM2 [17, 18] program, the brain of each patient was transformed by linear or nonlinear correlation to match the stan-dard brain. In eZIS analysis, a database of normal volunteers of various age groups which was built at the National Center Hospital for Mental, Nervous and Muscular Disorders, National Center of Neurology and Psychiatry, was applied. In the age-matched normal database after morphological transformation of the standard brain using the SPM2 program, the counts were normalized to the mean pixel values larger than one eighth of the total mean pixel value of all normal data, and a mean and SD image of each pixel was con-structed from the data. In a similar manner, patient data was also normalized with the total brain mean after morphological trans-formation to the standard brain using the SPM2 program. Then, the Z-value is calculated for each voxel using the following formu-la; Z-value = (mean pixel value in the normal group – pixel value in the case)/(standard deviation in the normal group). Consequently, the calculated mean Z-values were displayed on the anatomically standardized MRI template as the brain surface values.

vbSEE Application and Setting of ROIs After the acquisition of eZIS brain map, the vbSEE program was

applied for automatically converting the acquired data into the Ta-lairach brain coordinating space [19] . The vbSEE program pre-pared a distribution pattern of ROIs segmented into the Brodmann levels (level 5), in which the obtained brain coordinates correspond to anatomical information in accordance with Talairach Daemon [20] , and conducted Z-value association between the coordinates in the prepared patterns of ROIs and the case coordinates [21] . Con-sequently, all the coordinates of each subject provided Z-value in-formation. Subsequently, the Z-value data of 47 ROIs in every Brod-mann area (BA) of the left hemisphere were obtained according to segments and the anatomical classification of Talairach Daemon ( fig. 1 ). In this study, BA 1, 2, 3, 4, 6, 24, and 32 were selected for studied areas based on a finding in the literature which suggested the crucial roles of these areas in the swallowing process [4–6] .

Data Collection and Statistical Analysis To clarify the contrast of rCBF between damaged and acti-

vated cortical areas, the vbSEE (which enables to separately ex-press relative rCBF decrease and increase in each ROI) was ap-plied [22] . To clearly show a significant rCBF reduction in each region in the left hemisphere, we used the extent value, which can be calculated using the following formula: extent (%) = (number of coordinates which Z-value was more than 2 in the ROI/total number of coordinates in the ROI) ! 100 [9, 23] . As a statistical analysis, clinical characteristics and extent value in each ROI were compared between two groups with a logistic regression test and � 2 test with JMP software (SAS Institute Inc.). For ROIs which showed a significant difference in extent value between two groups, the receiver operating characteristic (ROC) curves were determined using the ROCKIT 0.9 program [24] . The program calculates the area under the ROC curves (Az ), accuracy, sensitiv-ity, and specificity for detection of dysphagia. In addition, inter-polated values for drawing ROC curves was automatically calcu-lated using the PlotROC program [25] .

Results

The clinical characteristics of studied patients in both groups are shown in table 1 . There was no signifi-cant difference in the characteristics between the dys-phagia and control groups, except for the level of FOIS. The level of FOIS was significantly higher in the control group than in the dysphagia group (p ! 0.001), confirm-ing the difference of swallowing ability between two groups. We have summarized lesions of CTs in one ad-ditive synthesis figure of all the patients with dysphagia and the control group ( fig. 2 ). Lesions did not extend to BA 4 or BA 24 in either group. Table 2 shows the extent value of each ROI in the two groups and the p value of a logistic regression test. A significant difference in the

Fig. 1. ROI on brain surface images and axial images anatomically classified Talairach Daemon level 5 (Brod-mann level) on vbSEE.

Co

lor v

ersi

on

avai

lab

le o

nlin

e

Dow

nloa

ded

by:

Uni

vers

ity o

f Tok

yo

13

3.11

.204

.54

- 11

/15/

2014

9:2

3:04

AM

Cortical Areas Causing Dysphagia after Stroke

Eur Neurol 2012;67:74–80 77

extent value between the two groups was found in BA 4 (p = 0.006) and BA 24 (p = 0.008). For these two ROIs, ROC analysis was performed and the results of the anal-ysis are shown in figure 3 . For BA 4, the Az value was 0.89 with a sensitivity of 80% and specificity of 60% for detecting the presence of dysphagia, respectively. The optimal cutoff value for the discrimination in the area was 13.8% of the extent value. Similarly, the Az value in BA 24 was 0.69 with a sensitivity of 60% and specificity of 60% for detecting the presence of dysphagia, respec-tively. The optimal cutoff value was 32.1% of the extent value in the area. Figure 4 represents BA 4 and BA 24. A representative case of a 70-year-old male with dysphagia is shown in figure 5 .

Discussion

We were able to reveal the significant difference in rCBF on SPECT in some cortical areas between post-stroke patients with and without dysphagia using the eZIS program and vbSEE program for SPECT analysis. To our knowledge, this is the first report showing the as-sociation of rCBF changes on SPECT in cortical areas and the development of dysphagia as a result of SPECT imag-ing analyzed with some recently developed novel ana-lytic methods.

Some authors have reported the significant role of cor-tical lesions such as BA 1, 2, 3, 4, 6, 24, and 32 for the de-velopment of dysphagia, although this association has not been emphasized previously. Larson et al. [26] report-ed that the destruction of the precentral gyrus impairs oral stage of swallowing in monkeys. Similarly, Narita

Table 2. Comparison of extent value (Z-value >2.0)

Left brain Dysphagia (n = 10) Controls (n = 10) p

Hemisphere 35.9830.8 17.7820.7 0.112BA 1 17.0812.5 6.7813.4 0.183BA 2 28.7826.0 13.0813.4 0.084BA 3 25.8819.7 10.5814.7 0.053BA 4 30.5823.2 6.6812.2 0.006BA 6 23.4820.1 10.8814.0 0.097BA 24 42.4842.6 6.089.2 0.008BA 32 20.9815.0 18.6813.2 0.06

Table 1. Clinical characteristics of participants (n = 20)

Characteristics Dysphagia(n = 10)

Controls(n = 10)

p

Mean age (SD) 67.1 (4.2) 65.1 (5.9) 0.36Male, n (%) 8 (80) 6 (60) 0.33Lesion site, n

MCA areaThalamusPutamen

244

433

Hemorrhage count, n (%) 8 (80) 6 (60) 0.33Mean NIHSS (SD) 11.5 (1.4) 10.6 (2.1) 0.24FOIS (SD) 4 (1.1) 7 <0.001Onset to SPECT

days (SD) 42 (8.6) 39 (6.3) <0.48

M CA = Middle cerebral artery; NIHSS = NIH stroke scale.

Fig. 2. Summary of CT lesions.

Co

lor v

ersi

on

avai

lab

le o

nlin

e

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

False positive fraction

True

pos

itiv

e fr

acti

on

BA4: Az = 0.89 Cutoff value: 13.8% Sensitivity: 80% Specificity: 60%

BA24: Az = 0.69 Cutoff value: 32.1%Sensitivity: 60% Specificity: 60%

Fig. 3. Conventional binormal ROC curves (extent) with data con-version to the core SPECT data for detection of dysphagia patients.

Dow

nloa

ded

by:

Uni

vers

ity o

f Tok

yo

13

3.11

.204

.54

- 11

/15/

2014

9:2

3:04

AM

Momosaki/Abo/Kakuda/Uruma

Eur Neurol 2012;67:74–8078

et al. [27] investigated whether motor cortex cooling in monkeys impairs swallowing reflex, and clarified that motor cortex does have an important role for induction and execution of swallowing.

Regarding the underlying mechanism of this cerebral cortical drive for swallowing, some investigations have already been made in humans. Hamdy et al. [4] revealed that the cingulate cortex, sensorimotor cortex, and supe-rior premotor cortex are activated by a swallowing move-ment with fMRI study. In addition, activation in the sen-sorimotor cortices and the cingulated and premotor cor-tex are produced by swallowing in the report by Mosier and Bereznaya [28] . In the report by Martin et al. [6] , sa-liva swallowing activated the anterior rostral cingulate cortex, and they found that the lateral precentral gyrus is deeply related to the programming of voluntary and water swallowing. These reports emphasized the impor-tance of the sensorimotor cortex for motor regulation and execution and sensorimotor control. In addition, it is speculated that the anterior rostral cingulate cortex plays a major role in higher order motor processing: swallowing movement planning and execution and cog-

nitive perceptual processes, such as attention and re-sponse selection. Our results showing the important role of BA 4 and BA 24 in the development of dysphagia are consistent with the results of these previous reports and speculation, although the previous studies did not apply novel analytic software as we did.

With SPECT evaluation using these novel analytic programs, the potential damage around stroke lesions can be expressed. This novel voxel-based analytic sys-tem is safe, computerized, and capable of accurate auto-matic and in-depth analysis. In this paper, we emphasize the superiority of SPECT evaluation analyzed by novel computerized methods for the diagnosis of dysphagia. The screening test by endoscopy is associated with the risk of aspiration and cannot be performed in some pa-tients since it is an invasive examination requiring the swallowing of real food. On the contrary, SPECT imag-ing can be safely performed even in unconscious pa-tients, demented patients with delirium, severely ill pa-tients, and in patients at high risk of aspiration pneumo-nia. In addition, the analysis can be performed by any operator (no need to be experienced). The results of the

Fig. 4. BA 4 and BA 24 defined by Talairach Daemon level 5 on vbSEE.

Dow

nloa

ded

by:

Uni

vers

ity o

f Tok

yo

13

3.11

.204

.54

- 11

/15/

2014

9:2

3:04

AM

Cortical Areas Causing Dysphagia after Stroke

Eur Neurol 2012;67:74–80 79

analysis hardly differ significantly among evaluations since it is automatically calculated. Furthermore, the cutoff values for the regions showing significant differ-ences were novel findings. The cutoff values could be determined because of the use of SPECT. Based on these values, the dysphagia-inducing level of injury could be determined. Therefore, it is recommended that SPECT analysis as we used should be performed to clarify the risk of dysphagia and aspiration for such patient popula-tions in which other screening tests for dysphasia are not suitable.

This study does have some limitations. First, the num-ber of cases was small. The findings of this study should be confirmed in a larger number of patients. Second, the subtype of stroke, the lesion size, and degree of dysphagia in the subjects were not homogenous. It is desirable to perform the analysis after the statistical corrections for

these clinical factors. Third, the period between the onset of stroke and SPECT imaging varied among the subjects. The extent of rCBF changes might have been influenced by the time-course after the onset of stroke.

Conclusions

Using the novel analytic software packages for SPECT, eZIS, and vbSEE, we were able to clarify the cortical areas related to the development of dysphagia in poststroke pa-tients safely. As a result, the identified areas are consistent with the previous reports. It is expected that this novel approach using SPECT for the evaluation of dysphagia might be useful for accurately determining the risk of dysphagia and consequent aspiration.

Fig. 5. CT and eZIS of a 70-year-old male with remarkable dysphagia caused by stroke in the left cortex. SPECT was performed 49 days after onset. Voxels of the Z-score 1 0 and extent 1 500 were colored as the parameter on a standard brain atlas.

Co

lor v

ersi

on

avai

lab

le o

nlin

e

Dow

nloa

ded

by:

Uni

vers

ity o

f Tok

yo

13

3.11

.204

.54

- 11

/15/

2014

9:2

3:04

AM

Momosaki/Abo/Kakuda/Uruma

Eur Neurol 2012;67:74–8080

References

1 Martino R, Foley N, Bhogal S, et al: Dyspha-gia after stroke: incidence, diagnosis, and pulmonary complications. Stroke 2005; 36: 2756–2763.

2 Gustafsson B, Tibbling L: Dysphagia, an unrecognized handicap. Dysphagia 1991; 6: 193–199.

3 Horner J, Buoyer FG, Alberts MJ, Helms MJ: Dysphagia following brain-stem stroke clin-ical correlates and outcome. Arch Neurol 1991; 48: 1170–1173.

4 Hamdy S, Mikulis DJ, Crawley A, Xue S, Lau H, Henry S, Diamant NE: Cortical activation during human volitional swallowing: an event-related fMRI study. Am J Physiol 1999; 277: 219–225.

5 Hamdy S, Rothwell JC, Brooks DJ, Bailey D, Aziz Q, Thompson DG: Identification of the cerebral loci processing human swallowing with H2(15)O PET activation. J Neurophysi-ol 1999; 81: 1917–1926.

6 Martin RE, Goodyear BG, Gati JS, Menon RS: Cerebral cortical representation of auto-matic and volitional swallowing in humans. J Neurophysiol 2001; 85: 938–950.

7 Mizumura S, Kumita S, Cho K, Ishihara M, Nakajo H, Toba M, et al: Development of quantitative analysis method for stereotactic brain image: assessment of reduced accumu-lation in severity and severity using anatom-ical segmentation. Ann Nucl Med 2003; 17: 289–295.

8 Van Laere KJ, Warwick J, Versijpt J, Goethals I, Audenaert K, Van Heerden B, Dierckx R: Analysis of clinical brain SPECT data based on anatomic standardization and reference to normal data: an ROC-based comparison of visual, semiquantitative, and voxel-based methods. J Nucl Med 2002; 43: 458–469.

9 Matsuda H, Mizumura S, Nagao T, Ota T, Iizuka T, Nemoto K, Takemura N, Arai H, Homma A: Automated discrimination be-tween very early Alzheimer disease and con-trols using an easy Z-score imaging system for multicenter brain perfusion single-pho-ton emission tomography. AJNR Am J Neu-roradiol 2007; 28: 731–736.

10 Katagi H, Uruma G, Uematsu M, Kobayashi K, Kakuda W, Abo M: Cerebral blood flow analyzed with the easy Z-score imaging sys-tem and voxel-based stereotactic extraction estimation in patients with unilateral dys-phagia (in Japanese). Jikeikai Med J 2008; 123: 237–247.

11 Wahlund L, Barkhof F, Fazekas F, Bronge L, Augustin M, Sjögren M: A new rating scale for age-related white matter changes appli-cable to MRI and CT. Stroke 2001; 32: 1318–1322.

12 Tohara H, Saitoh E, Mays KA, Kuhlemeier K, Palmar JB: Three tests for predicting aspira-tion without videofluorography. Dysphagia 2003; 18: 126–134.

13 Langmore SE, Schatz K, Olsen N: Fiberoptic endoscopic examination of swallowing safe-ty: a new procedure. Dysphagia 1988; 2: 216–219.

14 Colodny N: Interjudge and intrajudge reli-abilities in fiberoptic endoscopic evaluation of swallowing (FEES) using the penetration-aspiration scale: a replication study. Dyspha-gia 2002; 17: 308–315.

15 Crary MA, Mann GD, Groher ME: Initial psychometric assessment of a functional oral intake scale for dysphagia in stroke patients. Arch Phys Med Rehabil 2005; 86: 1516–1520.

16 Ohnishi T, Matsuda H, Hashimoto T, Kuni-hiro T, Nishikawa M, Uema T, et al: Abnor-mal regional cerebral blood flow in child-hood autism. Brain 2000; 123: 1838–1844.

17 Friston KJ, Ashburner J, Poline JB, Frith CD, Heather JD, Frackowiak RS: Spatial registra-tion and normalization of images. Hum Brain Mapp 1995; 3: 165–189.

18 Friston KJ, Frith CD, Fletcher P, Liddle PF, Frackowiak RS: Functional topography: multidimensional scaling and functional connectivity in the brain. Cereb Cortex 1996; 6: 156–164.

19 Talairach J, Tournoux P: Co-Planar Stereo-taxic Atlas of the Human Brain. New York, Thieme Medical Publishers, 1988.

20 Talairach Daemon. http://www.ihb.spb.ru/ � pet_lab/MSU/MSUMain.html (accessed July 29, 2009).

21 Mizumura S, Kumita S: Stereotactic statisti-cal imaging analysis of the brain using the easy Z-score imaging system for sharing a normal database. Radiat Med 2006; 24: 545–552.

22 Uruma G, Kakuda W, Abo M: Changes in regional cerebral blood flow in the right cor-tex homologous to left language areas are di-rectly affected by left hemispheric damage in aphasic stroke patients: evaluation by Tc-ECD SPECT and novel analytic software. Eur J Neuro 2010; 17: 461–469.

23 Matsuda H, Mizumura S, Soma T, Takemura N: Conversion of brain SPECT images be-tween different collimators and reconstruc-tion processes for analysis using statistical parametric mapping. Nucl Med Commun 2004; 25: 67–74.

24 Stephan C, Wesseling S, Schink T, Jung K: Comparison of eight computer programs for receiver-operating characteristic analysis. Clin Chem 2003; 49: 433–439.

25 Gorog G: An Excel program for calculating and plotting receiver-operator characteristic (ROC) curves, histograms and descriptive statistics. Comput Biol Med 1994; 24: 167–169.

26 Larson CR, Byrd KE, Garthwaite CR, Lush-iei ES: Alterations in the pattern of mastica-tion after ablation of the lateral precentral cortex in rhesus macaques. Exp Neurol 1980; 70: 638–651.

27 Narita N, Yamamura K, Yao D, Martin RE, Sessle BJ: Effects of functional disruption of lateral pericentral cerebral cortex on primate swallowing. Brain Res 1999; 824: 140–145.

28 Mosier K, Bereznaya I: Parallel cortical net-works for volitional control of swallowing in humans. Exp Brain Res 2001; 140: 280–289.

Dow

nloa

ded

by:

Uni

vers

ity o

f Tok

yo

13

3.11

.204

.54

- 11

/15/

2014

9:2

3:04

AM