J Neurol (1999) 246 : 939–942 © Steinkopff Verlag 1999 Kosuke Miura Yusaku Nakamura Fumiko Miura Ikuko Yamada Mitsuo Takahashi Akiteru Yoshikawa Toshiharu Mizobata Received: 27 November 1998 Received in revised form: 30 March 1999 Accepted: 26 April 1999 K. Miura (쾷) · Y. Nakamura · I. Yamada Department of Neurology, Sakai Hospital, Kinki University School of Medicine, 2-7-1, Harayamadai, Sakai, Osaka 590-0132, Japan Tel.: +81-722-991120, Fax: +81-722-986691 F. Miura · M. Takahashi Department of Neurology, Kinki University School of Medicine, 377-2, Ohno-Higashi, Osaka-Sayama, Osaka 598, Japan A. Yoshikawa · T. Mizobata Department of Radiology, Oriono-Izumi Hospital, Japan O R I G I N A L C O M M U N I C AT I O N Functional magnetic resonance imaging to word generation task in a patient with Broca’s aphasia Abstract We describe the findings of functional magnetic resonance imaging (fMRI) in a patient with Broca’s aphasia. The patient was a 45-year-old, right-handed woman who developed Broca’s aphasia after infarction in the left frontal lobe. The first fMRI showed no signals in the left frontal lobe during verbal tasks, 2 weeks after the onset of infarction. Four weeks later, when the patient’s symptom had improved, the second fMRI showed some increase in the fMRI signals in the left frontal lobe. Seven months later, she had completely recovered the ability to speak. The last fMRI then showed Introduction Since Ogawa et al. [8] reported the imaging of brain function using a relative decrease in deoxyhemoglobin (blood oxygenation level dependent) functional magnetic resonance imaging (fMRI) has been used to study the effect of various stimuli such as exercise, somatic stimulation, visual stimulation, auditory stimulation, speech, memory, and learning. There have been some studies of fMRI findings obtained with word-generation tasks in normal subjects [1, 2, 5, 7, 9] but not in patients with Broca’s aphasia. Our study was designed to study the fMRI findings in a patient with Broca’s aphasia. Patient and methods A Magnetom Vision 1.5-T system (Siemens, Erlangen, Germany) was used for fMRI. For anatomical positioning 15 axial slice im- that the increment in signal activity in the left frontal lobe during verbal tasks had recovered to the level seen in normal subjects. There was a good correlation between the increase in task-related signals in Broca’s area and the recovery of language function. Our findings show that fMRI has can be important in assessing cognitive functions in patients with Broca’s aphasia. Key words Functional magnetic resonance imaging · Cerebral infarction · Broca’s aphasia ages (3 mm thick, gapless) were obtained from the frontal to parietal regions by the T1-weighted spin echo method (repetition time/echo time = 500/15). From the T1-weighted images the slice planes containing the inferior frontal gyrus and insula were selected, and fMRI was performed using the fast low angle shot technique with a gradient echo sequence (repetition time/echo time/flip angle = 90 ms/56 ms/40, matrix = 128 × 256, slice thickness = 3 mm, field of view = 250 mm, acquisition time/image = 14 s). The verbal task used in the analysis was repeated vocalization of a stereotyped sentence “Tonari no Kyaku ha yoku Kaki kyu Kyaku da,” which is very familiar with almost of all Japanese as a tongue twister. It is not difficult for subjects to recall. The patient was asked to repeat it precisely, and as steady paced as possible. Each series consisted of five images obtained at rest and during the verbal task. Three series were obtained. The activated sites were evaluated with the following threestep procedure. Firstly, a functional map was produced from the images obtained using the Z-score [6]. To test for differences in signal between activation and rest the normal standard deviate, Z, was calculated for each pixel, as Z = (µA – µR)/SD, where µA is the mean signal intensity during activation, µR is the mean signal intensity during rest, and SD is the standard deviation of the measurements. In theory, activation would be considered significant if Z is larger than a given threshold. Secondly, the functional map 940 Table 1 Patient’s profile based on the SLTA: after 2 weeks expression in speaking and writing categories was more impaired than comprehension categories, after 2 months marked improvement was noted in the speaking and writing categories, and 7 months later the last SLTA showed full recovery to normal values 2 weeks 2 months 7 months Auditory comprehension Auditory word recognition Sentence comprehension Follow verbal commands Kana letter discrimination 65 100 90 20 50 100 100 100 100 100 100 100 100 100 100 Speaking Objective naming Word repetition Describe behaviors Explain picture story Sentence repetition Listing words (words fluency) Read aloud kanji words Read aloud kana letters Read aloud kana words Read aloud short sentence 55.3 55 100 40 0 60 8 100 90 100 0 97 100 100 100 80 100 90 100 100 100 100 99.5 100 100 100 100 100 95 100 100 100 100 was superimposed over the images of the same slices obtained by the T1-weighted spin echo method, and the activated sites were identified. Finally, at least two regions of interest (ROI) were established at the activated sites in the left inferior frontal gyrus and insula. In each ROI, a time-intensity curve was produced and evaluated. Changes were evaluated by this procedure 2 weeks, 4 weeks, and 7 months after the infarction. The Japanese Standard Language Test for Aphasia (SLTA) [4] was used to evaluate language functions. The SLTA is the standardized battery test most commonly used in Japan. Briefly, the correct responses in the SLTA testing are shown from 0% = most severe to 100% = normal. The following categories of the SLTA were analyzed in this study: auditory comprehension (to obey verbal commands), speaking (object naming, word repetition, sentence repetition), reading comprehension (follow written commands, sentence picture matching, etc.), writing (dictation of kana letters and dictation of short sentence), and calculation. Reading comprehension Kanji word-picture matching Kana word-picture matching Sentence picture matching Follow written commands 82.5 100 100 90 40 100 100 100 100 100 100 100 100 100 100 Writing Write kanji words Write kana words Narrative writing Dictate kana letters Dictate kanji words Dictate kana words Dictate short sentences 30 40 0 0 90 60 20 0 91.1 80 98 80 100 100 100 80 100 100 100 100 100 100 100 100 Calculation 50 90 98 Patient Unit: % Fig. 1 A, B MRI in a 45-year-old, right-handed woman with right hemiparesis and motor aphasia. A Spin-echo TR/TE: 500/15. B Spin-echo TR/TE: 4200/90. Her MRI reveals the presence of infarct lesions in the left inferior frontal gyrus, insula, white matter in the middle frontal gyrus, and the inferior area in the precentral gyrus A 45-year-old, right-handed woman noticed mild headache and difficulty in speech at about 6 : 00 a.m. on 28 May 1995, and was admitted to hospital. On admission, mild paresis was observed in the right upper and lower limbs. Seven days later, MRI showed infarct lesions in the left inferior frontal gyrus, insula, white matter in the middle frontal gyrus, and the inferior area in the precentral gyrus (Fig. 1). Magnetic resonance angiography, which was performed simultaneously, revealed a stenotic lesion in the middle cerebral artery M1 portion. On admission, neurological examination showed no abnormality in articulation of speech or the production of sound, but her spontaneous speech was not fluent. Writing short sentences to dictation was impossible. Sentence and word repetitions were less impaired than in short sentence dictation. Also, her handwriting was poor. On the other hand, her recognition was not so badly affected, although both auditory and reading comprehension was mildly impaired. In general, expression was more impaired than comprehension. The finding of the first examination of SLTA was compatible with Broca’s aphasia (Table 1). The time course of aphasia is shown in Table 1. Two months later, the second SLTA was performed; values had returned to normal except in some subset items of both speaking and writing cate- gories. Seven months later, the last SLTA showed improvement to the upper limit for almost all subsets. She gave informed consent, and the procedure for fMRI study was approved by the local ethics committee. Results fMRI performed 2 weeks after infarction showed no signals in response to the verbal tasks. Four weeks later, T1weighted spin echo MRI revealed lesions in the left frontal regions (Fig. 2 A). There were no definite signals in response to the verbal tasks in the infarct lesions, but there was an increase in blood flow in the left sylvian fissure (Fig. 2 B). Seven months later, T1-weighted MRI revealed old lesions in the left frontal regions (Fig. 3 A). Definite signals were elicited in response to the verbal 941 Fig. 2 A, B Four weeks later. A T1-weighted spin-echo imaging shows lesions in the left frontal regions. B fMRI shows no signals in response to word generation task in the infarct lesions but the increase in blood flow in the left sylvian fissure (arrow) Fig. 3 A, B Seven months later. A T1-weighted MRI shows old lesions in the left frontal regions. B fMRI shows definite signals with word generation task in the left inferior frontal gyrus, sylvian fissure, and precentral gyrus but no signals in the infarct lesions tasks in the left inferior frontal gyrus, sylvian fissure, and the precentral gyrus, but no signals appeared in the infarct lesions (Fig. 3 B). ROIs were established at the activated sites in the left inferior frontal gyrus. Figure 4 shows the time-intensity curve on the ROI. The signal in ROI significantly increased about 20% when the task was performed. This level was maintained during the verbal tasks but decreased after their discontinuation. Discussion fMRI can reveal the location of MRI signal changes associated with cerebral activity. There have been previous fMRI studies of word generation tasks in normal subjects; however, this technique has not been used to study word generation in patients with Broca’s aphasia. Broca’s area Fig. 4 ROI established at the activated sites in the left inferior frontal gyrus (arrow) in Fig. 3 B. The time-intensity curve on the ROI: signal increased by 20% with vocalization but decreased after the discontinuation of speech comprises the left inferior frontal gyrus (operuculum, trigone) and inferior part of the precentral gyrus [3]. Our patient’s clinical symptoms suggested Broca’s aphasia, and MRI showed infarct lesions in the inferior frontal gyrus, insular cortex, and precentral gyrus. fMRI performed 2 weeks later, when marked impairment of expression was observed, showed no signals with the verbal tasks. After 4 weeks, when repetition and comprehension had partially recovered, but she still could not recall the name of some materials, the increase in signals in the second fMRI study was not observed in the infarct lesions, but in the left sylvian fissure. Seven months later, when her clinical symptoms had improved, the last fMRI revealed an even greater increase in task-related signals with the verbal tasks in the left inferior frontal gyrus, sylvian fissure, and the precentral gyrus. Thus the fMRI findings showed functional recovery to the verbal tasks in the inferior frontal gyrus, which is the region responsible for speech, and also which reflected the course of recovery from Broca’s aphasia. Cuenod et al. [2] found that the fMRI signals with the word generation task was observed predominantly in the left cerebral hemisphere, mainly in the frontal lobe (Broca’s area). The fMRI findings in our patient after she had completely recovered from motor aphasia were consistent with those in their study. The mechanism of recovery from Broca’s aphasia may be reductions in brain edema in damaged tissues and the development of collateral circulation. Two categories of brain edema are recognized, vasogenic and cellular. In general, even when vasogenic edema improves, cellular edema persists for 3–4 weeks. Therefore the increase in brain blood flow in fMRI observed with the word generation task 4 weeks later suggested improved microcirculation due to reduced cellular edema. On the other hand, the greater increase in task-related signals in the left infe- 942 rior frontal gyrus 7 months later could represent functional recovery due to the development of collateral circulation. When Broca’s area (opercular and triangular portions of the left inferior frontal gyrus, or Brodmann’s areas 44, 45) is damaged, a transient or mild motor aphasia is observed [10]. Persistent Broca’s aphasia is thought to be due to damage not only in Broca’s area but also in more extensive areas containing the insula, frontal operculum, middle frontal gyrus, corona radiata, and arcuate fasciculus. However, these regional limits are not distinct. In our patient it was of interest to note that the increase in taskrelated signals in Broca’s area was associated with the recovery of language function. Our findings show that fMRI can be of importance in assessing cognitive functions in patients with Broca’s aphasia. References 1. Binder JR, Rao SM, Hammeke TA, Frost JA, Bandettini PA, Jesmanowicz A, Hyde JS (1995) Lateralized human brain language systems demonstrated by task subtraction functional magnetic resonance. Arch Neurol 52 : 593–601 2. Cuenod CA, Bookheimer SY, Pannier LH, Zeffiro TA, Theodore WH, LeBihan D (1995) Functional MRI during word generation, using conventional equipment. Neurology 45 : 1821–1827 3. Geschwind N, Putnam JJ (1980) L’aphasie de Broca. Le ph ix neurologique. Rev Neurol (Paris) 136 : 585–589 4. Hasegawa T (1977) Standard language test of aphasia manual of directions. Houmeidou, Tokyo pp 1–92 5. Hinke RM, Hu X, Stillman AE, Kim SG, Merkle H, Salmi R, Ugurbil K (1993) Functional magnetic resonance imaging of Broca’s area during internal speech. Neuroreport 4 : 675–678 6. Le Bihan D, Jazzard P, Turner R, Cuenod CA, Zeffiro TA (1993) Analysis of functional brain MR images with Z-maps (abstract). Magn Reson Imaging 3 : 141 7. McCarthy G, Blamire AM, Rothman DL, Gruetter R, Shulman RG (1993) Echo-planar magnetic resonance imaging studies of frontal cortex activation during word generation in humans. Proc Natl Acad Sci USA 90 : 4952– 4956 8. Ogawa S, Lee TM, Kay AR, Tank DW (1990) Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci USA 87 : 9868–9872 9. Rueckert L, Appollonio I, Grafman J, Jezzard P, Johnson R, Le Bihan D, Turner R, (1994) MRI functional activation of left frontal cortex during covert word production. J Neuroimaging 4 : 67–70 10. Tonkonogy J, Goodglass H (1981) Language function, foot of the third frontal gyrus, and rolandic operculum. Arch Neurol 38 : 486–490