HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 93/6/3687    most recent 01149.2004v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (16) Citing Articles via Google Scholar Google Scholar Articles by Tsujimoto, S. Articles by Sawaguchi, T. Search for Related Content PubMed PubMed Citation Articles by Tsujimoto, S. Articles by Sawaguchi, T.

REPORT

Neuronal Activity Representing Temporal Prediction of Reward in the Primate Prefrontal Cortex

¹Laboratory of Cognitive Neurobiology, Hokkaido University Graduate School of Medicine, Sapporo; and ²Core Research for Evolutional Science and Technology, Japan Science and Technology Agency, Saitama, Japan

Submitted 8 November 2004; accepted in final form 2 January 2005

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Temporal prediction of future events, especially regarding reward delivery, is critical for controlling/learning purposeful behavior. The dorsolateral prefrontal cortex (DLPFC) has been considered to be involved in behavioral control based on prospective coding for future events, including reward. Thus this area is likely to have a neuronal mechanism responsible for temporal prediction of forthcoming reward. To address this hypothesis, we recorded the neuronal activity from the DLPFC of macaque monkeys while they performed an oculomotor delayed-response task under two conditions regarding the time of reward delivery. In this task, when the subjects made a correct response, the reward was delivered after a reward-delay period of 0.5 or 2 s. At the behavioral level, the onset latency for saccades was significantly faster in the shorter reward-delay trials (0.5 s) than in longer reward-delay trials (2 s), indicating that our subjects actually predicted the time of reward delivery. At the neuronal level, we found that many DLPFC neurons showed differential activity depending on the predicted time of reward delivery during the cue and/or delay periods. These results suggest that a fraction of neurons in the DLPFC represent the temporal prediction of reward and probably a variety of other future events.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Temporal prediction of future events, particularly the time of reward (i.e., when the reward will be available), is critical for organizing/learning appropriate behaviors ( Dickinson et al. 1976; Hollerman and Schultz 1998). In primates, the dorsolateral prefrontal cortex (DLPFC) appears to be a candidate cortical region responsible for this cognitive ability. It has been repeatedly reported that activities of many DLPFC neurons are modulated by anticipated rewards ( Kobayashi et al. 2002; Leon and Shadlen 1999; Roesch et al. 2003; Wallis and Miller 2003; Watanabe 1996; Watanabe et al. 2002), suggesting that the DLPFC is involved in behavioral control based on prospective coding for future rewards, i.e., goal-directed behavior ( Hikosaka and Watanabe 2000; Schultz 2000). Furthermore, recent neuroimaging studies in humans have suggested that this area is involved in decision-making processes based on the expectancy of delays in reward delivery ( McClure et al. 2004; Tanaka et al. 2004). However, the neuronal basis of temporal prediction of reward delivery in the DLPFC remains to be examined.

In the present study, we hypothesized that the DLPFC contains a neuronal mechanism that represents the temporal prediction of reward. To test this hypothesis, we devised an oculomotor delayed-response (ODR) task under 2 conditions in which the subjects could predict when they would receive a reward after a correct response. We examined the neuronal activity recorded in the DLPFC of monkeys performing this task. We report here that a fraction of DLPFC neurons showed differential cue- and/or delay-period activity depending on the predicted time of reward delivery.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Two male macaque monkeys (Macaca mulatta, about 5.5 kg, and Macaca fuscata, about 7 kg, LN and SZ, respectively) served as subjects. Throughout this study, the subjects were treated in accordance with the "Guide for the Care and Use of Laboratory Animals" of the National Institutes of Health, and the experimental protocols were approved by the Animal Care and Use Committee of Hokkaido University School of Medicine.

The subjects were trained to perform a devised ODR task under 2 conditions regarding the time of reward delivery: short-delay (STD) and long-delay (LGD) conditions (Fig. 1A). Under both conditions, when the monkeys fixated on a central spot (a white square, 0.5° x 0.5°) for 1 s, a visual cue (a green cross or a red circle, 1.5° x 1.5°), which was associated with the time of reward (Fig. 1B), was presented at one of the 6 peripheral locations (eccentricity 15°, Fig. 1C) for 1 s. After a delay period of fixed duration (1.5 s), the fixation spot was turned off, which instructed the monkeys to make a memory-guided saccade to the cued location. When the eye movement fell inside a target window of 5°, a white square (0.5° x 0.5°) was presented at the cue position, and the reward-delay period of 0.5 or 2 s (STD or LGD conditions, respectively) began. In some trials, a small second saccade ensued to fixate on the target, but this eye movement did not result in differential neuronal activity between the 2 conditions during the cue and delay periods. When the monkey remained fixated during the reward-delay period, the same amount of liquid reward was delivered under both conditions. The 6 directions and 2 reward-delay conditions were ordered pseudo-randomly. Throughout the trial, the eye position was restricted to within 5° of the central fixation point or peripheral target. If the monkey broke fixation, the trial was aborted. Although we did not adopt the correction method, the number of fixation breaks was very small and the frequency was not different between the 2 reward-delay conditions.

View larger version (22K): [in this window] [in a new window]   FIG. 1. Behavioral task, recording sites, and behavioral results. A: oculomotor delayed-response (ODR) task in the present study. In this task, the subjects made a memory-guided saccade to a remembered target location that had been cued before a delay period of 1.5 s. A correct saccade was followed by a reward-delay period of 0.5 (short delay, STD) or 2 s (long delay, LGD), and then liquid reward was delivered. B: associations between stimulus and reward delay in each block of trials. C: central fixation spot and 6 peripheral possible locations of cue/target. D: schematic drawings of the recording sites. Recorded sites (shaded area) were in the dorsolateral prefrontal cortex rostral to the frontal eye field. E: scattergram of saccadic onset latencies; those in the STD trials are plotted against those in the LGD trials. Filled triangles and open circles represent the mean value for each recording session for monkeys LN and SZ, respectively. AS, arcuate sulcus; PS, principal sulcus.

The activity of single neurons was recorded using conventional electrophysiological techniques similar to those described in our previous studies ( Tsujimoto and Sawaguchi 2004a, b). Briefly, a glass-insulated elgiloy microelectrode (0.5–1.5 M) was positioned using a plastic grid with numerous small holes attached to the recording chamber. It was then advanced vertically using a pulse motor-driven micromanipulator (MO-81, Narishige, Tokyo, Japan) until the activity of one or more neurons could be recorded. Single-unit activities were isolated using a template-matching strategy with a multispike detector (Alpha Omega Engineering, Nazareth, Israel).

We focused on neurons in the DLPFC rostral to the frontal eye field (FEF) (Fig. 1D). To estimate the FEF physiologically, we applied intracortical microstimulation (ICMS; 22 cathodal pulses of 0.3-ms duration at 333 Hz, 100 µA) through the recording electrodes. When eye movements were elicited by the ICMS, the site was considered to be within the FEF ( Bruce et al. 1985) and data recorded from these sites were excluded from the study. After the recording was completed, the recording sites were confirmed using a standard histological technique described previously ( Iba and Sawaguchi 2002). The recorded neurons were located in the caudal half of area 46 and area 8a (Fig. 1D).

To examine the influence of the predicted time of reward delivery on DLPFC neuronal activity, we applied a 2-factor ANOVA (reward-delay conditions x directions; P < 0.05) on the mean firing rate during a cue and delay period, separately. We focused on the neurons whose cue- and/or delay-period activity showed a significant main effect in reward-delay conditions because these neurons appeared to be influenced by a temporal prediction of reward. To quantitatively examine the selectivity of the reward-delay conditions, we calculated a time selectivity index (TSI) as follows

where R_STD and R_LGD are the mean discharge rates during the cue or delay period for the direction with maximum activity for the STD and LGD conditions, respectively. This index ranges from –1 to +1. A value of 0 indicates identical responses to the 2 conditions, and positive and negative values indicate a preference for STD and LGD conditions, respectively. All analyses were performed on correct trials only.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Throughout the experimental sessions, both monkeys performed the task with >95% correct responses, and there were no significant differences in performance levels between the STD and LGD conditions. In >90% of the probe trials (see METHODS), both monkeys chose the target position that was associated with the short delay period (0.5 s). To compare the saccadic responses under the 2 conditions, we calculated the onset latencies of saccade from the go signal during each recording session, and plotted them on a scattergram (Fig. 1E). In almost all sessions, the onset latencies were faster in the STD trials than in the LGD trials, and the mean value was significantly different (STD vs. LGD, 213 ± 16 vs. 229 ± 16 ms, respectively, for monkey SZ; 227 ± 22 vs. 245 ± 19 ms, respectively, for monkey LN; 2-tailed Mann-Whitney U test, P < 0.01). These data indicate that our subjects predicted the time of reward delivery according to the visual stimuli.

We recorded activities from 347 DLPFC neurons. Of these, 121 neurons significantly modulated their activity during the cue and/or delay periods depending on the duration of the reward delays (n = 47 for cue period only, n = 56 for delay period only, and n = 18 for both cue and delay periods). As summarized in Table 1, 30 of 47 (64%) neurons that exhibited differential activity only during the cue period and 20 of 56 (36%) neurons that exhibited differential activity only during the delay period showed a significantly higher discharge rate in the STD trials than in the LGD trials. Of the 18 neurons that exhibited differential activity during both the cue and delay periods, 67% showed higher cue-period activity in the STD trials than in the LGD trials and 50% showed higher delay-period activity in the STD trials than in the LGD trials. In most of these neurons (69% for cue and 62% for delay), the discharge rate during the cue period and/or delay period was significantly different across the 6 directions of the target.

View larger version (51K): [in this window] [in a new window]   FIG. 2. Example of dorsolateral prefrontal cortex (DLPFC) neurons in which delay-period activity differed according to predicted time of reward. A: raster displays and averaged histograms, aligned by the cue (C), delay (D), and go signal (G), are shown separately for 6 directions and 2 reward-delay conditions. Top and bottom rows: STD and LGD conditions, respectively, and each column is associated with the direction indicated above them. Colored areas show the delay period. Red and blue colors represent the neuron's preferred and nonpreferred time, respectively. B: polar plots of the delay-period activity shown in A. Percentage changes in delay-period activity compared with the control period of 0.5 s that immediately preceded the cue presentation are plotted for each direction and each condition. C: averaged histograms of STD and LGD trials of the neuron in A for the direction with maximal activity (180°) are shown together. Bin width is 40 ms. Each curve is smoothed by averaging the 3 adjacent bins. D: mean discharge rate during the delay period for the direction with maximum activity (180°) of the neuron in A is shown separately for each task block. During the recording of this neuron, the block was reversed twice, i.e., this neuron was recorded in a total of 3 blocks (2 reversed blocks and one original block). Visual cues superimposed on each bar represent the stimulus associated with the reward delay. Error bars show ±SE.

Figure 3A illustrates an example of neurons that showed different cue-period activity according to the difference in predicted time of reward. Specifically, this neuron showed clear phasic activation during the cue period, which was spatially tuned especially for the upper right (60°) and upper left (120°) directions [2-way ANOVA, F(5,271) = 42.06, P < 0.001]. Furthermore, the magnitude of this cue-period activity was significantly greater under STD than under LGD conditions [2-way ANOVA, F(1,271) = 10.49, P < 0.01]. As shown in Fig. 3B, this difference in cue-period activity (STD > LGD) was reproduced when the block/association between cue and time was reversed.

View larger version (27K): [in this window] [in a new window]   FIG. 3. A: example of DLPFC neurons in which cue-period activity differed according to the predicted time of reward. Colored areas show the cue period. Other format and abbreviations are the same as those in Fig. 2A. B: mean discharge rate during the cue period for the direction with maximum activity (120°) of the neuron in A are shown separately for each task block. Error bars show ±SE.

View larger version (24K): [in this window] [in a new window]   FIG. 4. Population activity. A: each neuron's value for the time selectivity index (TSI) during the delay period was plotted against that during the cue period. Each plot shows the TSI values of individual neurons. Open squares and small circles represent neurons with and without statistically significant differences, respectively. Filled triangle and a filled rhombus represent the neurons shown in Figs. 2A and 3A, respectively. B and C: population activities normalized to a 0.5-s control period for DLPFC neurons with different cue-period (B) and delay-period activities (C) are shown. Red and blue lines show the population activity for the preferred and nonpreferred conditions of each neuron (not of the subjects), respectively. Bin width is 40 ms, and is smoothed by averaging across 3 adjacent bins. Error bars show ±SE.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
In the present study, we recorded the neuronal activity from the DLPFC of monkeys while they performed an ODR task in which the subjects predicted when the reward would be available. We found that the cue-period and/or delay-period activity of many DLPFC neurons was modulated by the different durations of forthcoming reward delay. These differential activities were not ascribed to the physical properties of the visual cues because the neuronal preference for the time was reproduced even when the time-cue association was reversed. Because the percentage of correct responses did not differ under the 2 conditions, the probability that the subject could get the reward was the same under the 2 conditions. These findings suggest that a fraction of neurons in the DLPFC represent temporal prediction of reward (i.e., when the reward will be delivered). Furthermore, considering the previous finding that the activity of many DLPFC neurons encodes several kinds of future information, including forthcoming actions ( Asaad et al. 1998; Hasegawa et al. 1998), anticipated visual objects ( Rainer et al. 1999), and rewards ( Kobayashi et al. 2002; Leon and Shadlen 1999; Roesch et al. 2003; Wallis and Miller 2003; Watanabe 1996), it is likely that DLPFC neurons represent temporal prediction not only of reward delivery, but also of a variety of other future events, although this hypothesis remains to be examined in further studies.

Previous studies have reported that task-related neuronal activity in the DLPFC is modulated by expectancy of the type or amount of reward ( Kobayashi et al. 2002; Leon and Shadlen 1999; Roesch et al. 2003; Wallis and Miller 2003; Watanabe 1996). These neurons have been considered to play a role in reward-based response selection/decision-making processes ( Schultz 1998, 2000). The present data introduce a new dimension that modulates the activities of DLPFC neurons (i.e., the time of reward), which is also a critical variable in such cognitive processes ( Dickinson et al. 1976; Hollerman and Schultz 1998; McClure et al. 2004; Tanaka et al. 2004). These data are consistent with the recent findings from human neuroimaging studies that suggest that the DLPFC is involved in decision-making processes based on the expectation of delays until the delivery of monetary reward ( McClure et al. 2004). The DLPFC neurons found here would be the neuronal substrate that contributes to behavioral control based on the prediction of time of reward.

Alternatively, the differentially responsive neurons identified here may encode not the time, but the subjective value of the reward. A complete dissociation of these 2 factors is difficult, and we do not exclude the subjective-value interpretation. In our task paradigm, however, the value of the reward was determined only by the temporal information, and our subjects received extensive overtraining on this task before and during neuronal recording. Furthermore, psychological studies have reported that a key component of predictions in learning is the time of reinforcement ( Dickinson et al. 1976), and many efficient theoretical models of learning are based on this observation ( Schultz 1998). Moreover, neurons in the DLPFC have been shown to encode a variety of temporal information, such as temporal order ( Ninokura et al. 2003) and duration ( Sakurai et al. 2004). Therefore we prefer the temporal-prediction interpretation rather than the subjective-value interpretation, although this problem should be examined in subsequent research by manipulating both the time and magnitude of the reward.

Adaptive behavior in dynamically changing environments requires not only expectation of future events, but also evaluation of the outcomes ( Matsumoto and Tanaka 2004). To predict reward delivery temporally would be important in a comparison of the expected reward and the received reward ( Dickinson et al. 1976; Hollerman and Schultz 1998), and the time-discriminating DLPFC neurons identified here may play a major role in this function. Interestingly, a fraction of DLPFC neurons was shown to encode the response-outcome of an immediately preceding action ( Tsujimoto and Sawaguchi 2004a, 2005), which may be a neuronal correlate of evaluating an immediate response-outcome. Considering these data together, the DLPFC appears to play a major role in both response selection based on the prediction of future events, including its timing, and in the evaluation of the actual outcome, thereby allowing the appropriate adaptation to dynamically changing environments.

	GRANTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
This work was supported by the Japan Society for the Promotion of Science (15009927) to S. Tsujimoto, the Ministry of Education, Culture, Sports, Science and Technology of Japan to T. Sawaguchi, and the Japan Science and Technology Agency.

	ACKNOWLEDGMENTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
We thank E. Ishida and K. Watanabe-Sawaguchi for assistance with animal care and surgery.

	FOOTNOTES

 
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Address for reprint requests and other correspondence: T. Sawaguchi, Laboratory of Cognitive Neurobiology, Hokkaido University Graduate School of Medicine, N15W7, Kita-ku, Sapporo 060-8638, Japan (E-mail: toshi-sw{at}med.hokudai.ac.jp)

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Asaad WF, Rainer G, and Miller EK. Neural activity in the primate prefrontal cortex during associative learning. Neuron 21: 1399–1407, 1998.[CrossRef][ISI][Medline]

Bruce CJ, Goldberg ME, Bushnell MC, and Stanton GB. Primate frontal eye fields. II. Physiological and anatomical correlates of electrically evoked eye movements. J Neurophysiol 53: 606–635, 1985.

Dickinson A, Hall G, and Mackintosh NJ. Surprise and the attenuation of blocking. J Exp Psychol Anim Behav Process 2: 313–322, 1976.[CrossRef]

Hasegawa R, Sawaguchi T, and Kubota K. Monkey prefrontal neuronal activity coding the forthcoming saccade in an oculomotor delayed matching-to-sample task. J Neurophysiol 79: 322–333, 1998.[Abstract/Free Full Text]

Hikosaka K and Watanabe M. Delay activity of orbital and lateral prefrontal neurons of the monkey varying with different rewards. Cereb Cortex 10: 263–271, 2000.[Abstract/Free Full Text]

Hollerman JR and Schultz W. Dopamine neurons report an error in the temporal prediction of reward during learning. Nat Neurosci 1: 304–309, 1998.[CrossRef][ISI][Medline]

Iba M and Sawaguchi T. Neuronal activity representing visuospatial mnemonic processes associated with target selection in the monkey dorsolateral prefrontal cortex. Neurosci Res 43: 9–22, 2002.[ISI][Medline]

Kobayashi S, Lauwereyns J, Koizumi M, Sakagami M, and Hikosaka O. Influence of reward expectation on visuospatial processing in macaque lateral prefrontal cortex. J Neurophysiol 87: 1488–1498, 2002.[Abstract/Free Full Text]

Leon MI and Shadlen MN. Effect of expected reward magnitude on the response of neurons in the dorsolateral prefrontal cortex of the macaque. Neuron 24: 415–425, 1999.[CrossRef][ISI][Medline]

Matsumoto K and Tanaka K. The role of the medial prefrontal cortex in achieving goals. Curr Opin Neurobiol 14: 178–185, 2004.[CrossRef][ISI][Medline]

McClure SM, Laibson DI, Loewenstein G, and Cohen JD. Separate neural systems value immediate and delayed monetary rewards. Science 306: 503–507, 2004.[Abstract/Free Full Text]

Ninokura Y, Mushiake H, and Tanji J. Representation of the temporal order of visual objects in the primate lateral prefrontal cortex. J Neurophysiol 89: 2868–2873, 2003.[Abstract/Free Full Text]

Rainer G, Rao SC, and Miller EK. Prospective coding for objects in primate prefrontal cortex. J Neurosci 19: 5493–5505, 1999.[Abstract/Free Full Text]

Roesch MR and Olson CR. Impact of expected reward on neuronal activity in prefrontal cortex, frontal and supplementary eye fields and premotor cortex. J Neurophysiol 90: 1766–1789, 2003.[Abstract/Free Full Text]

Sakurai Y, Takahashi S, and Inoue M. Stimulus duration in working memory is represented by neuronal activity in the monkey prefrontal cortex. Eur J Neurosci 20: 1069–1080, 2004.[CrossRef][ISI][Medline]

Schultz W. Predictive reward signal of dopamine neurons. J Neurophysiol 80: 1–27, 1998.[Abstract/Free Full Text]

Schultz W. Multiple reward signals in the brain. Nat Rev Neurosci 1: 199–207, 2000.[ISI][Medline]

Tanaka SC, Doya K, Okada G, Ueda K, Okamoto Y, and Yamawaki S. Prediction of immediate and future rewards differentially recruits cortico-basal ganglia loops. Nat Neurosci 7: 887–893, 2004.[CrossRef][ISI][Medline]

Tsujimoto S and Sawaguchi T. Neuronal representation of response–outcome in the primate prefrontal cortex. Cereb Cortex 14: 47–55, 2004a.[Abstract/Free Full Text]

Tsujimoto S and Sawaguchi T. Properties of delay-period neuronal activity in the primate prefrontal cortex during memory- and sensory-guided saccade tasks. Eur J Neurosci 19: 447–457, 2004b.[CrossRef][ISI][Medline]

Tsujimoto S and Sawaguchi T. Context-dependent representation of response–outcome in monkey prefrontal neurons. Cereb Cortex 2005 In press.

Wallis JD and Miller EK. Neuronal activity in primate dorsolateral and orbital prefrontal cortex during performance of a reward preference task. Eur J Neurosci 18: 1–13, 2003.[CrossRef][ISI][Medline]

Watanabe M. Reward expectancy in primate prefrontal neurons. Nature 382: 629–632, 1996.[CrossRef][Medline]

Watanabe M, Hikosaka K, Sakagami M, and Shirakawa S. Coding and monitoring of motivational context in the primate prefrontal cortex. J Neurosci 22: 2391–2400, 2002.[Abstract/Free Full Text]

This article has been cited by other articles:

				A. Correa and A. C. Nobre Neural Modulation by Regularity and Passage of Time J Neurophysiol, September 1, 2008; 100(3): 1649 - 1655. [Abstract] [Full Text] [PDF]

				J. Tanji and E. Hoshi Role of the Lateral Prefrontal Cortex in Executive Behavioral Control Physiol Rev, January 1, 2008; 88(1): 37 - 57. [Abstract] [Full Text] [PDF]

				J.-W. Sohn and D. Lee Order-Dependent Modulation of Directional Signals in the Supplementary and Presupplementary Motor Areas J. Neurosci., December 12, 2007; 27(50): 13655 - 13666. [Abstract] [Full Text] [PDF]

				T. Ikeda and O. Hikosaka Positive and Negative Modulation of Motor Response in Primate Superior Colliculus by Reward Expectation J Neurophysiol, December 1, 2007; 98(6): 3163 - 3170. [Abstract] [Full Text] [PDF]

				A. Genovesio, S. Tsujimoto, and S. P. Wise Neuronal Activity Related to Elapsed Time in Prefrontal Cortex J Neurophysiol, May 1, 2006; 95(5): 3281 - 3285. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 93/6/3687    most recent 01149.2004v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (16) Citing Articles via Google Scholar Google Scholar Articles by Tsujimoto, S. Articles by Sawaguchi, T. Search for Related Content PubMed PubMed Citation Articles by Tsujimoto, S. Articles by Sawaguchi, T.