用近红外功能成像技术研究语义和知觉编码过程中左前额叶的活动
作者：李鹏程 龚辉 杨炯炯 曾绍群 骆清铭 管林初
单位：李鹏程.华中理工大学生物医学光子学研究所，武汉 430074
关键词：近红外光谱；语义编码；知觉编码；左前额叶活动
航天医学与医学工程000201摘要：目的 用近红外光学功能成像技术研究语义与知觉编码过程中的左前额叶活动。方法 实时记录22名青年被试者(男女各半)在语义和知觉加工过程中760 nm和850 nm两个波长近红外连续光透过额骨和左前额皮层后的漫射光强变化，并以此表征该皮层区血容(CBV)和血氧的变化。语义加工要求被试者用两个非相关的汉字词对组成一个有意义的句子；知觉加工则要求被试者判断两个词是否具有相同的结构组合以及两个词对中左右结构的差别。 结果 结果显示，与知觉加工相比，语义加工时左下前额叶活动增多。该区域大致对应布德曼45、47区的一部分。 结论 成功地利用近红外功能成像技术观测了语义与知觉编码过程中的左前额叶活动，获得了与fMRI和断层成象技术(PET)研究相近的结果，进一步证实了左前额叶在语义加工中的作用。
, 百拇医药
中图分类号：R319；R338.2 文献标识码：A 文章编号：1002-0837(2000)02-0079-05
Left Prefrontal Cortex Activation during Semantic Encoding Accessed with Functional Near Infrared Imaging
LI Peng-cheng,GONG Hui,YANG Jiong-jiong,ZENG Shao-qun
(Institute of Biomedical Photonics, Huazhong University of Science and Technology, Wuhan 430074)
Abstract: Objective To investigate the left prefrontal lobe activation during semantic and non-semantic encoding tasks with functional near-infrared imaging (fNIRI) technique. Method 22 healthy subjects were assigned semantic encoding and non-semantic encoding tasks. During semantic encoding tasks, subjects were asked to make a meaningful sentence including two unrelated Chinese word pairs, while during non-semantic encoding task they were asked to judge whether the two Chinese word pairs had the same morphological structure or not. Light intensity of two wavelengths (760 nm and 850 nm) diffused through skull and left prefrontal lobe were real-time recorded and used to reconstruct the brain activation image during the experiment. Result With the fNIRI, significant activations were observed in the left inferior prefrontal cortex (Brodmann' areas 45 and 47) during the two tasks, but the evoked activations were more significant for semantic than non-semantic task. These observations were consistent with the results reported by others with functional megnetic resonance imaging(fMRI) and positron-emission tomographyPET.Conclusion The results suggest that fNIRI provides an important, non-invasive way to map the prefrontal activation during cognitive tasks.
, 百拇医药
Key word:near-infrared spectroscopy;semantic encoding;non-semanticencoding;left prefrontal activation
Modern functional neuroimaging methods, such as positron-emission tomography (PET), functional MRI (fMRI) and optical imaging with NIRS, utilize activity-dependent hemodynamic changes to obtain indirect maps of the evoked electrical activity during brain activation. These imaging techniques rely on various types of activity-dependent hemodynamic changes: Regional changes in cerebral blood flow (CBF) were used for functional mapping of the brain by PET,single photon emission computed tomography (SPECT) and flow-sensitive MRI. The optical imaging technique relies on the optical properties of cortex change due to regional changes of either cerebral blood oxygenation or cerebral blood volume (CBV) changes when the cortex is functionally active. Since near-infrared light can penetrate deeply into tissues, optical method employing near-infrared radiation could detect the absorption or scattering changes of the cortex through skin and skull.
, 百拇医药
Since Petersen et al[1] provided the first functional neuroimaging evidence accessed with PET implicating left prefrontal regions in the semantic analysis of words, subsequent PET and fMRI studies have found similar left prefrontal activations indicating LIPC involving the semantic encoding[2].In these studies, the left inferior prefrontal cortex (LIPC) area, anterior to Broca's area, showed greater activation during semantic than non-semanticencoding[2,3].As a unique non-invasive technique for brain function imaging, near-infrared spectroscopy has been used to study brain activities, including motor[4], visual[5] and cognitive[6] activities. However, this technique has not been used in studying the brain activation during semantic encoding tasks. In this paper, left prefrontal activation during encoding unrelated word pairs were measured by means of functional near-infrared imaging (fNIRI) and difference between activations during semantic encoding tasks and non-semanticencoding tasks were demonstrated.
, 百拇医药
Methods
Subjects The subjects of the experiment were 22 right-handed healthy students of Huazhong University of Science and Technology (1/2 were female, aged 18 to 22 years). All subjects are native speakers of Chinese and had no history of neurological or psychiatric disorders.
Materials The stimuli were taken from 50 unrelated Chinese word pairs with no obvious semantic relation between the two words in each word pair. The difficulty and structure of these word pairs are carefully selected and matched in difficulty and frequency. Of the 50 word pairs, 40 pairs were used as material for experiment. The remaining 10 pairs were used for exercising and filling. The 40 experimental word pairs were divided into 2 groups. Each group contained 20 word pairs used for semantic or non-semanticencoding session. Word pairs in each group were presented to the subjects in a random order by a computer.
, http://www.100md.com
Procedure The procedure of the experiment included:(a) Rest condition (silent, baseline): The subjects were instructed to stare at a “+" character at the center of the screen and keep as calm as possiblei;(b) Semantic encoding task: The subjects were asked to make a meaningful sentence including the presented word pair on the screen and speak it out;(c) Non-semantic encoding task: The subjects were asked to judge whether the two words in the presented word-pairs had the same morphological structure or not. If not, figure out the difference in LEFT-RIGHT structure( LEFT-RIGHT structure is one of the 5 morphological way to construct a Chinese character).
, 百拇医药
The experiment sequence for each subject is, baseline (30 s), task (150 s), rest (60 s), the other task (150 s), and rest (60 s). The order of altering encoding tasks was counterbalanced across subjects. Each word pair was presented centrally on the computer screen for 5 s, followed by the character “+" presented in a second interval before the appearance of the next word pairs. Word pairs orders were kept randomly across subjects.
Principle of functional Near-Infrared Imaging The fNIRI used in this study is designed basing on the dual wavelength (760 nm and 850 nm) near-infrared spectrophotometry[7,8].It is capable of picking up small difference in absorption characteristics between the oxygenated and deoxygenated states of the hemoglobin at 760 nm and 850 nm. Utilizing these two wavelengths, it is possible to determine the relative concentrations of oxygenated (HbO2), deoxygenated (Hb) hemoglobin and blood volume between the two states[9]. Since the neural activity is associated with hemodynamic changes, it can be accessed by fNIRI. The detail description of fNIRI was given in reference [8].
, 百拇医药
Experiment setup The diagram of functional near-infrared imager used in this study is shown as figure 1. Nine Tungsten lamps and four dual wavelength detectors make 16 lamp-detector pairs covering an area of 9 cm by 4 cm. In each detection section, the distance between lamp and detector pair is 2.5 cm. Each dual wavelength detector pair has two silicon photodiodes (9 mm by 11 mm), which are protected by optical filters at 760 nm or 850 nm (with half width 9 nm), respectively. Diffused photons through the skull and cortex under these sections were detected and transformed to represent the relative concentration changes of oxygenation of hemoglobin and blood volume in the measured region. In the experiments, the imager was put on the surface of the left prefrontal area, with the right reach to the middle of the forehead, and the upper close to the hairline, as shown in figure 1. Signals from these 16 channels were sampled and encoded to calculate the △O.D. of blood volume. In order to make data analysis more clarified, the 16 channels of the imager probe were divided into four subregions: A, B, C, D, as shown in figure 2.
, 百拇医药
Fig.1 Diagram of fNIRI (functional near-infrared imager) for left prefrontal region
Fig.2 Sketch of channels and subregions of imager
Time-resolution for each channel is 0.25 s and each scan takes 4 s when run on a 133 MHz personal computer. The time-resolution is limited by the velocity of the computer and the rise time of the light source which is the main factor.
, http://www.100md.com
Data analysis Basing on a modified Beer-Lambert's law[8],we used △O.Dbv to represent the relative concentration changesof CBV. If △O.Dbv >0, the measured region was considered activated,otherwise,not activated. △O.Dbv was calculated by
△O.Dbv=△O.D850nm+k. △O.D760nm
△O.D.=log(I0/Itask)-log(I0/Irest)
, http://www.100md.com
=log(Irest)-log(Itask)
where I0 represents the incident light intensity,Itask and Irest represent the light intensities detected by the fNIRI during task and rest period respectively.k is the modification factor for reducing the crosstalk between changes of blood volume and oxygenation. This modification factor is determined by calibration on a blood model.
For each subject, the △O.Dbv were averaged in each period (e.g. baseline, task and recovery etc.). Subtraction between the △O.Dbv of task and that of the baseline or rest would indicate the activation (localization and response intensity) of the task relative to baseline or rest. Subtraction between semantic and non-semanticencoding task would show activation of semantic encoding task relative to non-semanticencoding task. To obtain activation over all the subsjects, responses were averaged over all subjects.
, 百拇医药
Result
During the experiment,17 subjects completed the semantic encoding task and 18 subjects completed the non-semanticencoding task.
Figure 3 displays two typical results of diffused light intensity changes of 850 nm recorded by fNIRI in the experiment. The curve was connected through B-Spline fitting. As shown in figure 3, a decrease of diffusion light intensity was presented whil esubjects made the tasks compared to the rest condition, which suggested a CBV increase in relevant region during the tasks. Though blood volume increases were observed during both tasks, they were of different amplitude. The amplitude was generally higher during semantic encoding than during non-semanticencoding, which implicated that the corresponding region was more activated during semantic encoding.
, 百拇医药
Fig.3 Typical curve of diffused light intensity change via time during the tasks
Table 1 Percentage of activation in different subregions during the two tasks task
subregion
activated
nonactivated
percentage of activated
A
4
, 百拇医药
18
0.19
semantic
B
5
17
0.23
encoding
C
10
12
0.45
D
, 百拇医药
8
14
0.36
A
6
16
0.27
nonsemantic
B
6
16
0.27
encoding
, http://www.100md.com
C
7
15
0.32
D
10
12
0.45
Table 2 Percentage of activation of subtracting semantic with nonsemantic subregion
actiated
nonactivated
, http://www.100md.com
percentage of activated
A
4
14
0.29
B
1
17
0.06
C
12
6
0.67
, 百拇医药
D
9
9
0.50
This CBV increase was observed in all 22 subjects, but the pattern varied with individuals. Though the activation patterns were different with different individuals, the localization of the activation showed significant convergence across all the subjects. Therefore, the regional distribution of the peak activation is also of interest. Statistic analysis was done on activated proportion under semantic encoding and non-semanticencoding tasks in comparison with rest condition in different regions. As shown in table 1, there was a higher activation percentage in C and D than in A and B during both semantic encoding and non-semanticencoding tasks (χ2=360, 85.4,P<0.005). Comparing the activated proportion during semantic encoding tasks with that during non-semanticencoding tasks(table 2), the results showed that the difference among regions was significant and region C (corresponding to inferior frontal cortex,part of Brodman's areas 45 and 47) was activated more than region D (χ2=360,P<0.01).
, 百拇医药
Table 3a and 3b showed the average of blood volume in 4 subregions in response to semantic encoding and non-semanticencoding tasks respectively. Subtraction between a and b was put in table 3c, which showed greater averaged fNIRI activation during semantic encoding relative to non-semanticencoding and the area of peak response is clearly focused on region C.Table 3 Average △O.Dbv in different subregions during two tasks subregion
average △O.Dbv
, 百拇医药
a
b
c
A
0.19
0.009
0.010
B
0.028
0.014
0.014
C
0.057
, http://www.100md.com
0.020
0.038
D
0.026
0.033
0.008
Discussion
In the case of evoked response studies, the hemoglobin concentration changes are assumed in the brain only. In fact, the fNIRI signal comes not only from the cortex, but also from the skin, skull and CSF(cere-brospinal fluid). Therefore, changes happened there during different periods of the experiment would affect the results. In the present experiment, we used SUBTRACTION between those during the task and during rest (or baseline) to remove this kind of contribution.
, http://www.100md.com
Activation of the LIPC was observed in both semantic encoding and non-semanticencoding tasks.The activation is evoked more during semantic encoding than during non-semantic encoding. Despite that some variabilities were observed among the results from PET and fMRI studies of activations in LIPC during semantic encoding, there have been many agreements[1,3].Our data showed strong convergence about the activation of brain areas. This also provides additional support for the claim that noninvasive optical recording provides a simple tool for brain function study.Optical method can be performed in adults under conditions of simulated or real stress that may have important effects upon functional performance, or where the subject can not be well immobilized. Considering its low cost and non-invasiveness, these features make fNIRI a promising new tool for the study of the dynamics of brain metabolic activity during cognitive and behavioral tasks.
, http://www.100md.com
Foundation item:Supported by National Nature Science Foundation of China(39770228)
骆清铭，邓晖，龚 辉等. 用于脑血流检测的近红外光谱术[J].红外与毫米波学报，1999，18(2):138～144
作者地址：李鹏程.华中理工大学生物医学光子学研究所，武汉 430074
[References]
[1] Petersen SE, Fox PT, Posner MI et al.Positron emission tomographic studies of cortical anatomy of single-word processing [J]. Nature,1988, 331: 585～589
, 百拇医药
[2] Gabrieli JE, Poldrack RA, Deamond JE. The role of left prefrontal cortex in language and memory [J]. Proc Natl Acad Sci USA, 1998, 95:906～913
[3] Demb JB, Desmond JE, Wagner AD et al. Semantic encoding and retrieval in the left inferior prefrontal cortex: a fNIRI study of task difficulty and process specificity [J]. The Journal of Neuroscience,1995,15(9): 5870～5878
[4] Villringer A.Monitoring of cerebral blood oxygenation during human brain activation by simultaneous high-resolution MRI and near-infrared spectroscopy[C]. Proceeding of the Society of Magnetic Resonance San Francisco California 1994:67
, http://www.100md.com
[5] Hoshi Y，Tamua M. Dynamic multichannel near-infrared optical imaging of human brain activity [J]. J Appl Physiol，1993, 75: 1842～1846
[6] Chance B, Zhuang Z,Unah C et al.Cognition—activated low—frequency modulation of light absorption in human brain[J]. Proc Natl Acad Sci USA,1993, 90:3770～3774
[7] Chance B, Luo QM, Nioka S. Optical investigations of physiology: a study of intrinsic and extrinsic biomedical contrast[J]. Phil Trans R Soc. London: B, 1997, 352: 707～716
[8] Luo Q, Nioka S, Chance B.Functional near-infrared imager[C], SPIE, 1997, 2979: 84～93
[9] LUO Qingming,DENG Hui,GONG Hui et al.New-infrared spectroscopy for the measurement of cerebval blood flow[J].Journal of Infrarel and Miuimeter Waves,1999,18(2):138～144
Received date:1999-07-21, http://www.100md.com
单位：李鹏程.华中理工大学生物医学光子学研究所，武汉 430074
关键词：近红外光谱；语义编码；知觉编码；左前额叶活动
航天医学与医学工程000201摘要：目的 用近红外光学功能成像技术研究语义与知觉编码过程中的左前额叶活动。方法 实时记录22名青年被试者(男女各半)在语义和知觉加工过程中760 nm和850 nm两个波长近红外连续光透过额骨和左前额皮层后的漫射光强变化，并以此表征该皮层区血容(CBV)和血氧的变化。语义加工要求被试者用两个非相关的汉字词对组成一个有意义的句子；知觉加工则要求被试者判断两个词是否具有相同的结构组合以及两个词对中左右结构的差别。 结果 结果显示，与知觉加工相比，语义加工时左下前额叶活动增多。该区域大致对应布德曼45、47区的一部分。 结论 成功地利用近红外功能成像技术观测了语义与知觉编码过程中的左前额叶活动，获得了与fMRI和断层成象技术(PET)研究相近的结果，进一步证实了左前额叶在语义加工中的作用。
, 百拇医药
中图分类号：R319；R338.2 文献标识码：A 文章编号：1002-0837(2000)02-0079-05
Left Prefrontal Cortex Activation during Semantic Encoding Accessed with Functional Near Infrared Imaging
LI Peng-cheng,GONG Hui,YANG Jiong-jiong,ZENG Shao-qun
(Institute of Biomedical Photonics, Huazhong University of Science and Technology, Wuhan 430074)
Abstract: Objective To investigate the left prefrontal lobe activation during semantic and non-semantic encoding tasks with functional near-infrared imaging (fNIRI) technique. Method 22 healthy subjects were assigned semantic encoding and non-semantic encoding tasks. During semantic encoding tasks, subjects were asked to make a meaningful sentence including two unrelated Chinese word pairs, while during non-semantic encoding task they were asked to judge whether the two Chinese word pairs had the same morphological structure or not. Light intensity of two wavelengths (760 nm and 850 nm) diffused through skull and left prefrontal lobe were real-time recorded and used to reconstruct the brain activation image during the experiment. Result With the fNIRI, significant activations were observed in the left inferior prefrontal cortex (Brodmann' areas 45 and 47) during the two tasks, but the evoked activations were more significant for semantic than non-semantic task. These observations were consistent with the results reported by others with functional megnetic resonance imaging(fMRI) and positron-emission tomographyPET.Conclusion The results suggest that fNIRI provides an important, non-invasive way to map the prefrontal activation during cognitive tasks.
, 百拇医药
Key word:near-infrared spectroscopy;semantic encoding;non-semanticencoding;left prefrontal activation
Modern functional neuroimaging methods, such as positron-emission tomography (PET), functional MRI (fMRI) and optical imaging with NIRS, utilize activity-dependent hemodynamic changes to obtain indirect maps of the evoked electrical activity during brain activation. These imaging techniques rely on various types of activity-dependent hemodynamic changes: Regional changes in cerebral blood flow (CBF) were used for functional mapping of the brain by PET,single photon emission computed tomography (SPECT) and flow-sensitive MRI. The optical imaging technique relies on the optical properties of cortex change due to regional changes of either cerebral blood oxygenation or cerebral blood volume (CBV) changes when the cortex is functionally active. Since near-infrared light can penetrate deeply into tissues, optical method employing near-infrared radiation could detect the absorption or scattering changes of the cortex through skin and skull.
, 百拇医药
Since Petersen et al[1] provided the first functional neuroimaging evidence accessed with PET implicating left prefrontal regions in the semantic analysis of words, subsequent PET and fMRI studies have found similar left prefrontal activations indicating LIPC involving the semantic encoding[2].In these studies, the left inferior prefrontal cortex (LIPC) area, anterior to Broca's area, showed greater activation during semantic than non-semanticencoding[2,3].As a unique non-invasive technique for brain function imaging, near-infrared spectroscopy has been used to study brain activities, including motor[4], visual[5] and cognitive[6] activities. However, this technique has not been used in studying the brain activation during semantic encoding tasks. In this paper, left prefrontal activation during encoding unrelated word pairs were measured by means of functional near-infrared imaging (fNIRI) and difference between activations during semantic encoding tasks and non-semanticencoding tasks were demonstrated.
, 百拇医药
Methods
Subjects The subjects of the experiment were 22 right-handed healthy students of Huazhong University of Science and Technology (1/2 were female, aged 18 to 22 years). All subjects are native speakers of Chinese and had no history of neurological or psychiatric disorders.
Materials The stimuli were taken from 50 unrelated Chinese word pairs with no obvious semantic relation between the two words in each word pair. The difficulty and structure of these word pairs are carefully selected and matched in difficulty and frequency. Of the 50 word pairs, 40 pairs were used as material for experiment. The remaining 10 pairs were used for exercising and filling. The 40 experimental word pairs were divided into 2 groups. Each group contained 20 word pairs used for semantic or non-semanticencoding session. Word pairs in each group were presented to the subjects in a random order by a computer.
, http://www.100md.com
Procedure The procedure of the experiment included:(a) Rest condition (silent, baseline): The subjects were instructed to stare at a “+" character at the center of the screen and keep as calm as possiblei;(b) Semantic encoding task: The subjects were asked to make a meaningful sentence including the presented word pair on the screen and speak it out;(c) Non-semantic encoding task: The subjects were asked to judge whether the two words in the presented word-pairs had the same morphological structure or not. If not, figure out the difference in LEFT-RIGHT structure( LEFT-RIGHT structure is one of the 5 morphological way to construct a Chinese character).
, 百拇医药
The experiment sequence for each subject is, baseline (30 s), task (150 s), rest (60 s), the other task (150 s), and rest (60 s). The order of altering encoding tasks was counterbalanced across subjects. Each word pair was presented centrally on the computer screen for 5 s, followed by the character “+" presented in a second interval before the appearance of the next word pairs. Word pairs orders were kept randomly across subjects.
Principle of functional Near-Infrared Imaging The fNIRI used in this study is designed basing on the dual wavelength (760 nm and 850 nm) near-infrared spectrophotometry[7,8].It is capable of picking up small difference in absorption characteristics between the oxygenated and deoxygenated states of the hemoglobin at 760 nm and 850 nm. Utilizing these two wavelengths, it is possible to determine the relative concentrations of oxygenated (HbO2), deoxygenated (Hb) hemoglobin and blood volume between the two states[9]. Since the neural activity is associated with hemodynamic changes, it can be accessed by fNIRI. The detail description of fNIRI was given in reference [8].
, 百拇医药
Experiment setup The diagram of functional near-infrared imager used in this study is shown as figure 1. Nine Tungsten lamps and four dual wavelength detectors make 16 lamp-detector pairs covering an area of 9 cm by 4 cm. In each detection section, the distance between lamp and detector pair is 2.5 cm. Each dual wavelength detector pair has two silicon photodiodes (9 mm by 11 mm), which are protected by optical filters at 760 nm or 850 nm (with half width 9 nm), respectively. Diffused photons through the skull and cortex under these sections were detected and transformed to represent the relative concentration changes of oxygenation of hemoglobin and blood volume in the measured region. In the experiments, the imager was put on the surface of the left prefrontal area, with the right reach to the middle of the forehead, and the upper close to the hairline, as shown in figure 1. Signals from these 16 channels were sampled and encoded to calculate the △O.D. of blood volume. In order to make data analysis more clarified, the 16 channels of the imager probe were divided into four subregions: A, B, C, D, as shown in figure 2.
, 百拇医药
Fig.1 Diagram of fNIRI (functional near-infrared imager) for left prefrontal region
Fig.2 Sketch of channels and subregions of imager
Time-resolution for each channel is 0.25 s and each scan takes 4 s when run on a 133 MHz personal computer. The time-resolution is limited by the velocity of the computer and the rise time of the light source which is the main factor.
, http://www.100md.com
Data analysis Basing on a modified Beer-Lambert's law[8],we used △O.Dbv to represent the relative concentration changesof CBV. If △O.Dbv >0, the measured region was considered activated,otherwise,not activated. △O.Dbv was calculated by
△O.Dbv=△O.D850nm+k. △O.D760nm
△O.D.=log(I0/Itask)-log(I0/Irest)
, http://www.100md.com
=log(Irest)-log(Itask)
where I0 represents the incident light intensity,Itask and Irest represent the light intensities detected by the fNIRI during task and rest period respectively.k is the modification factor for reducing the crosstalk between changes of blood volume and oxygenation. This modification factor is determined by calibration on a blood model.
For each subject, the △O.Dbv were averaged in each period (e.g. baseline, task and recovery etc.). Subtraction between the △O.Dbv of task and that of the baseline or rest would indicate the activation (localization and response intensity) of the task relative to baseline or rest. Subtraction between semantic and non-semanticencoding task would show activation of semantic encoding task relative to non-semanticencoding task. To obtain activation over all the subsjects, responses were averaged over all subjects.
, 百拇医药
Result
During the experiment,17 subjects completed the semantic encoding task and 18 subjects completed the non-semanticencoding task.
Figure 3 displays two typical results of diffused light intensity changes of 850 nm recorded by fNIRI in the experiment. The curve was connected through B-Spline fitting. As shown in figure 3, a decrease of diffusion light intensity was presented whil esubjects made the tasks compared to the rest condition, which suggested a CBV increase in relevant region during the tasks. Though blood volume increases were observed during both tasks, they were of different amplitude. The amplitude was generally higher during semantic encoding than during non-semanticencoding, which implicated that the corresponding region was more activated during semantic encoding.
, 百拇医药
Fig.3 Typical curve of diffused light intensity change via time during the tasks
Table 1 Percentage of activation in different subregions during the two tasks task
subregion
activated
nonactivated
percentage of activated
A
4
, 百拇医药
18
0.19
semantic
B
5
17
0.23
encoding
C
10
12
0.45
D
, 百拇医药
8
14
0.36
A
6
16
0.27
nonsemantic
B
6
16
0.27
encoding
, http://www.100md.com
C
7
15
0.32
D
10
12
0.45
Table 2 Percentage of activation of subtracting semantic with nonsemantic subregion
actiated
nonactivated
, http://www.100md.com
percentage of activated
A
4
14
0.29
B
1
17
0.06
C
12
6
0.67
, 百拇医药
D
9
9
0.50
This CBV increase was observed in all 22 subjects, but the pattern varied with individuals. Though the activation patterns were different with different individuals, the localization of the activation showed significant convergence across all the subjects. Therefore, the regional distribution of the peak activation is also of interest. Statistic analysis was done on activated proportion under semantic encoding and non-semanticencoding tasks in comparison with rest condition in different regions. As shown in table 1, there was a higher activation percentage in C and D than in A and B during both semantic encoding and non-semanticencoding tasks (χ2=360, 85.4,P<0.005). Comparing the activated proportion during semantic encoding tasks with that during non-semanticencoding tasks(table 2), the results showed that the difference among regions was significant and region C (corresponding to inferior frontal cortex,part of Brodman's areas 45 and 47) was activated more than region D (χ2=360,P<0.01).
, 百拇医药
Table 3a and 3b showed the average of blood volume in 4 subregions in response to semantic encoding and non-semanticencoding tasks respectively. Subtraction between a and b was put in table 3c, which showed greater averaged fNIRI activation during semantic encoding relative to non-semanticencoding and the area of peak response is clearly focused on region C.Table 3 Average △O.Dbv in different subregions during two tasks subregion
average △O.Dbv
, 百拇医药
a
b
c
A
0.19
0.009
0.010
B
0.028
0.014
0.014
C
0.057
, http://www.100md.com
0.020
0.038
D
0.026
0.033
0.008
Discussion
In the case of evoked response studies, the hemoglobin concentration changes are assumed in the brain only. In fact, the fNIRI signal comes not only from the cortex, but also from the skin, skull and CSF(cere-brospinal fluid). Therefore, changes happened there during different periods of the experiment would affect the results. In the present experiment, we used SUBTRACTION between those during the task and during rest (or baseline) to remove this kind of contribution.
, http://www.100md.com
Activation of the LIPC was observed in both semantic encoding and non-semanticencoding tasks.The activation is evoked more during semantic encoding than during non-semantic encoding. Despite that some variabilities were observed among the results from PET and fMRI studies of activations in LIPC during semantic encoding, there have been many agreements[1,3].Our data showed strong convergence about the activation of brain areas. This also provides additional support for the claim that noninvasive optical recording provides a simple tool for brain function study.Optical method can be performed in adults under conditions of simulated or real stress that may have important effects upon functional performance, or where the subject can not be well immobilized. Considering its low cost and non-invasiveness, these features make fNIRI a promising new tool for the study of the dynamics of brain metabolic activity during cognitive and behavioral tasks.
, http://www.100md.com
Foundation item:Supported by National Nature Science Foundation of China(39770228)
骆清铭，邓晖，龚 辉等. 用于脑血流检测的近红外光谱术[J].红外与毫米波学报，1999，18(2):138～144
作者地址：李鹏程.华中理工大学生物医学光子学研究所，武汉 430074
[References]
[1] Petersen SE, Fox PT, Posner MI et al.Positron emission tomographic studies of cortical anatomy of single-word processing [J]. Nature,1988, 331: 585～589
, 百拇医药
[2] Gabrieli JE, Poldrack RA, Deamond JE. The role of left prefrontal cortex in language and memory [J]. Proc Natl Acad Sci USA, 1998, 95:906～913
[3] Demb JB, Desmond JE, Wagner AD et al. Semantic encoding and retrieval in the left inferior prefrontal cortex: a fNIRI study of task difficulty and process specificity [J]. The Journal of Neuroscience,1995,15(9): 5870～5878
[4] Villringer A.Monitoring of cerebral blood oxygenation during human brain activation by simultaneous high-resolution MRI and near-infrared spectroscopy[C]. Proceeding of the Society of Magnetic Resonance San Francisco California 1994:67
, http://www.100md.com
[5] Hoshi Y，Tamua M. Dynamic multichannel near-infrared optical imaging of human brain activity [J]. J Appl Physiol，1993, 75: 1842～1846
[6] Chance B, Zhuang Z,Unah C et al.Cognition—activated low—frequency modulation of light absorption in human brain[J]. Proc Natl Acad Sci USA,1993, 90:3770～3774
[7] Chance B, Luo QM, Nioka S. Optical investigations of physiology: a study of intrinsic and extrinsic biomedical contrast[J]. Phil Trans R Soc. London: B, 1997, 352: 707～716
[8] Luo Q, Nioka S, Chance B.Functional near-infrared imager[C], SPIE, 1997, 2979: 84～93
[9] LUO Qingming,DENG Hui,GONG Hui et al.New-infrared spectroscopy for the measurement of cerebval blood flow[J].Journal of Infrarel and Miuimeter Waves,1999,18(2):138～144
Received date:1999-07-21, http://www.100md.com