부호화된 사건의 시간적 정보를 기반으로 한 인출은 일화기억의 중요한 통제기제 중 하나이다. 기억인출과 관련한 수많은 신경영상 연구들이 진행되었음에도 아직 시간적으로 구성된 일화기억의 인출에 관여하는 뇌신경연결망 패턴에 대해서는 알려진 바가 많지 않다. 본 연구에서는 두가지 다른 순차적 인출 뇌신경 기제를 구분하기 위하여 과제기반 기능적 연결성 다변량패턴분석 방법을 사용하였다. 참가자들은 시간적 일화기억과제를 수행하였고, 순서대로 부호화된 기억자극을 순방향 혹은 역방향으로 인출하도록 지시를 받았다. 부분적으로 분류된 국소적 신경네트워크 패턴은 두 인출기제를 잘 구분하지 못한 반면, 기억과 관련된 인지통제 영역과 목표-지향적 인지기제처리에 관련된 것으로 알려진 여러 피질-피질하 노드들을 아우르는 전뇌신경네트워크 패턴은 시간적 일화기억 인출기제를 잘 구분하였다. 이 영역들은 측면/내측 전전두엽 영역, 하부 두정엽, 중간 측두회, 선조체 영역 등을 포함하며 기계학습을 이용한 분류에서 높은 분류 예측률을 보였다. 본 연구의 결과는 일화기억의 시간적 인출기제에 관여하는 피질-피질하 여러 영역의 관여를 확인하였고, 대역적 네트워크 패턴의 기능적 연결성이 질적으로 다른 인출기제에 관여함을 확인하였다는데에 중요성을 갖는다.
부호화된 사건의 시간적 정보를 기반으로 한 인출은 일화기억의 중요한 통제기제 중 하나이다. 기억인출과 관련한 수많은 신경영상 연구들이 진행되었음에도 아직 시간적으로 구성된 일화기억의 인출에 관여하는 뇌신경연결망 패턴에 대해서는 알려진 바가 많지 않다. 본 연구에서는 두가지 다른 순차적 인출 뇌신경 기제를 구분하기 위하여 과제기반 기능적 연결성 다변량 패턴분석 방법을 사용하였다. 참가자들은 시간적 일화기억과제를 수행하였고, 순서대로 부호화된 기억자극을 순방향 혹은 역방향으로 인출하도록 지시를 받았다. 부분적으로 분류된 국소적 신경네트워크 패턴은 두 인출기제를 잘 구분하지 못한 반면, 기억과 관련된 인지통제 영역과 목표-지향적 인지기제처리에 관련된 것으로 알려진 여러 피질-피질하 노드들을 아우르는 전뇌신경네트워크 패턴은 시간적 일화기억 인출기제를 잘 구분하였다. 이 영역들은 측면/내측 전전두엽 영역, 하부 두정엽, 중간 측두회, 선조체 영역 등을 포함하며 기계학습을 이용한 분류에서 높은 분류 예측률을 보였다. 본 연구의 결과는 일화기억의 시간적 인출기제에 관여하는 피질-피질하 여러 영역의 관여를 확인하였고, 대역적 네트워크 패턴의 기능적 연결성이 질적으로 다른 인출기제에 관여함을 확인하였다는데에 중요성을 갖는다.
Retrieving temporal information of encoded events is one of the core control processes in episodic memory. Despite much prior neuroimaging research on episodic retrieval, little is known about how large-scale connectivity patterns are involved in the retrieval of sequentially organized episodes. Tas...
Retrieving temporal information of encoded events is one of the core control processes in episodic memory. Despite much prior neuroimaging research on episodic retrieval, little is known about how large-scale connectivity patterns are involved in the retrieval of sequentially organized episodes. Task-related functional connectivity multivariate pattern analysis was used to distinguish the different sequential retrieval. In this study, participants performed temporal episodic memory tasks in which they were required to retrieve the encoded items in either the forward or backward direction. While separately parsed local networks did not yield substantial efficiency in classification performance, the large-scale patterns of interactivity across the cortical and sub-cortical brain regions implicated in both the cognitive control of memory and goal-directed cognitive processes encompassing lateral and medial prefrontal regions, inferior parietal lobules, middle temporal gyrus, and caudate yielded high discriminative power in classification of temporal retrieval processes. These findings demonstrate that mnemonic control processes across cortical and subcortical regions are recruited to re-experience temporally-linked series of memoranda in episodic memory and are mirrored in the qualitatively distinct global network patterns of functional connectivity.
Retrieving temporal information of encoded events is one of the core control processes in episodic memory. Despite much prior neuroimaging research on episodic retrieval, little is known about how large-scale connectivity patterns are involved in the retrieval of sequentially organized episodes. Task-related functional connectivity multivariate pattern analysis was used to distinguish the different sequential retrieval. In this study, participants performed temporal episodic memory tasks in which they were required to retrieve the encoded items in either the forward or backward direction. While separately parsed local networks did not yield substantial efficiency in classification performance, the large-scale patterns of interactivity across the cortical and sub-cortical brain regions implicated in both the cognitive control of memory and goal-directed cognitive processes encompassing lateral and medial prefrontal regions, inferior parietal lobules, middle temporal gyrus, and caudate yielded high discriminative power in classification of temporal retrieval processes. These findings demonstrate that mnemonic control processes across cortical and subcortical regions are recruited to re-experience temporally-linked series of memoranda in episodic memory and are mirrored in the qualitatively distinct global network patterns of functional connectivity.
* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.
문제 정의
The aim of the current study was to elucidate patterns of functional connectivity associated with mnemonic control processes that were implicated in backward or forward retrieval of sequentially organized episodic events. We demonstrated that patterns of interactivity across a large-scale brain network constructed distinct functional connectivity structures during the two different directional retrieval conditions.
제안 방법
After completing the temporal memory task, participants performed a working memory task, N-back, during which they were successively presented with a series of single-alphabet letters. The working memory task is employed to ensure that the expected sequential memory activation does not merely result from the working memory load difference but rather from the demand required for controlled retrieval processes.
Functional images were then realigned to the first volume for motion correction, spatially normalized to the Montreal Neurological Institute (MNI) template provided with SPM8, then resampled into 3 × 3 × 3 mm3 size voxels, followed by spatial smoothing using a Gaussian kernel with a full width at a half maximum of 8 mm.
In order to avoid any biases resulting from the node selection based on task-activated patterns such as GLM findings, the nodes for additional functional connectivity analyses were defined based on the results of the intrinsic resting-state analysis. The seed for resting-state connectivity analysis was defined within the aVLPFC region exclusively implicated in the temporal retrieval rather than working memory process in our data (see Results section 3.
Next, the current experiment also specifically aimed to collectively elucidate the whole-brain pattern characteristics of networks and to address how these are related to the regions previously implicated in controlled retrieval regions. We compared the achieved pattern classification performance across separately parsed networks to reveal whether our discriminating features (i.
If the former account is correct, we would expect increased activation of mid-VLPFC, which has been implicated in domain-general selection processes that follow retrieval and resolve competition among many retrieved representations. Our results demonstrated the increased involvement of left aVLPFC during backward retrieval supporting the latter account consistent with controlled retrieval processes explanation.
Informed consent was obtained in a manner approved by the Institutional Review Board of Yonsei University. Prior to the experiment, participants were screened for any significant medical conditions, including their history of neurological or psychiatric diagnoses.
To obtain task-related time-series reflecting psychological factors, BOLD signal time-series were extracted from the whole brain in a voxel-wise manner for all retrieval runs using in-house scripts in MATLAB. The extracted BOLD signal time-series were then deconvolved with a parametric empirical Bayesian formulation implemented in SPM to estimate voxel-by-voxel neural time-series, and these were detrended to remove any linear and nonlinear trends. Condition-specific psychological factors were defined by contrasting the onset time points of each retrieval condition against baseline (i.
Given that pattern analysis is more sensitive to extract subtle differences from distributed multi-dimensional functional effect than is GLM, we employed multivariate pattern analysis approach based on the machine learning algorithm. The novel approach adopted in our methods was that we extracted neural time-series across the whole retrieval runs and then convolved them with psychological factors (i.e., two different retrieval directions) to generate condition-weighted time-series in a voxel-by-voxel manner and take the cross-correlation between all voxels into account. This multivariate approach also has advantages in that the researchers do not need to pre-define ROIs (i.
Therefore, this method resulted in a total of 465 accuracy measures. The purpose of this approach was not to find the algorithm that maximized the prediction accuracy. Instead, it was designed to explore which features were potentially informative in classifying different retrieval types.
This study employed the fcMVPA approach to elucidate brain networks informative in distinguishing the backward retrieval process from the forward retrieval process. The results revealed that the large-scale functional connectivity patterns encompassing brain regions previously implicated in controlled retrieval process and goal-directed cognitive process, were substantially effective in decoding the sequential retrieval processes (i.
6). To examine how these intrinsically connected brain regions (i.e., 31 nodes, Table 3 and Fig. 4) interacted with each other during temporal episodic memory task, we conducted fcMVPA using task-based time-series extracted from the retrieval run. Specifically, we used fcMVPA to investigate whether forward and backward retrieval processes recruited qualitatively different patterns of functional connectivity.
5. To investigate the number of features (i.e., functional connectivity between two different brain regions) required to produce optimally informative patterns in a classification algorithm, we performed a linear SVM pattern classification which iterated as a function of the number of features included. The SVM classification accuracies were 60% with only one feature and gradually increased to 87.
To understand the relative importance of each brain region in transferring information across the network, we computed the normalized betweenness centrality of each node included in the functional connectivity patterns that conferred peak classification accuracy. Betweenness centrality represents a node’s centrality in a certain network and is computed as the total number of shortest paths between each pair of all other nodes that passes through the node (Freeman, 1977, 1978; Girvan & Newman, 2002).
, 2012a, 2012b), we gradually increased the number of features included with iteration of SVM classification processes, which resulted in a series of classification accuracies. With this protocol, we could monitor the changes in the pattern of accuracy as a function of the number of features included; and as a result, we were able to detect the features of functional connectivity that constructed the most informative patterns in distinguishing between the two different retrieval conditions.
대상 데이터
Each SVM classification performance was estimated using a leave-one-subject-out cross validation procedure. For each iteration of the validation, one subject (one sample from the forward and another sample from the backward types) was removed and the remaining 19 subjects (thus 19 forward and 19 backward types) were put into the machine learning algorithm to train the classifier. The excluded two samples from one subject were then used to test the performance of the classifier: if the classifier predicted the class label of each test sample correctly, accuracy was scored as “1,” whereas if an incorrect prediction was made, it was scored as “0.
Neuroimaging data were acquired with the 3T General Electric Healthcare Discovery MR750 (Waukesha, WI) using an 8-channel radiofrequency head coil. Functional data were obtained with a T2*-weighted gradient-echo echoplanar imaging sequence (TR = 2000 ms, TE = 30 ms, 3.
The temporal episodic memory task consisted of an encoding and a retrieval runs. The experimental items of the episodic memory task were composed of 90 and 160 common Korean nouns (2 syllable words with 2 Korean letters long). All word items were projected onto a screen with a black background and viewed using a mirror mounted on the head coil.
Twenty-two healthy volunteers (8 females; mean age = 23.23, SD = 2.39) participated in the experiment and were paid for their time (10,000 South Korean Won/hour). Two participants were excluded from all analyses because of incompletion due to fatigue, and the analysis for the N-back task was conducted with data from only 19 participants because data for one participant were unavailable.
데이터처리
, 31 × 30/2) were used as features for further connectivity analyses. For feature selection, t-tests using correlation coefficients from two retrieval directions were performed for each feature comparing the differences between the means of the two conditions. Then, 465 features were ranked based on the absolute t-score values to eliminate any directional bias, and z-score transformation was applied to each feature (r) to improve normality.
, 2002). The least-square parameter estimates of the best-fitting synthetic HRF for each condition of interest were used in pair-wise contrasts and stored as separate images for each participant; these were then checked against the null hypothesis with one-tailed t-tests to determine whether effects of participants were random at the group level. Given that the primary aim of the current study is to conduct fcMVPA, GLM analyses adopted less stringent uncorrected threshold to explore sufficient number of potential node regions.
1). The time series of seed region was extracted and was correlated with the time courses of all other voxels using the Pearson cross-correlation. The intrinsic connectivity map obtained from the resting-state analysis generated 42 peak regions based on SPM-identified whole sub-peak local maxima.
이론/모형
Each SVM classification performance was estimated using a leave-one-subject-out cross validation procedure. For each iteration of the validation, one subject (one sample from the forward and another sample from the backward types) was removed and the remaining 19 subjects (thus 19 forward and 19 backward types) were put into the machine learning algorithm to train the classifier.
성능/효과
08 Hz). Furthermore, nuisance covariates including six head motion parameters, global mean signal, white matter signal, and cerebrospinal fluid signal were regressed out to improve the validity of our findings. Seed-to-voxel connectivity maps were generated for each participant and then were entered into the group level analysis.
The classification procedures occurred as conducted with the original data. If the connectivity patterns for the two retrieval directions were truly different and if each type had distinctive combinations of connectivity, randomly shuffled class labels for SVM classification should result in an accuracy of approximately 50%, whereas classification based on the original class labels should yield accuracy above chance.
The excluded two samples from one subject were then used to test the performance of the classifier: if the classifier predicted the class label of each test sample correctly, accuracy was scored as “1,” whereas if an incorrect prediction was made, it was scored as “0.” Therefore, there were 40 accuracies estimated from 20 rounds of iterations, which were then averaged to obtain one representative accuracy measure for the Nth SVM classification.
This study employed the fcMVPA approach to elucidate brain networks informative in distinguishing the backward retrieval process from the forward retrieval process. The results revealed that the large-scale functional connectivity patterns encompassing brain regions previously implicated in controlled retrieval process and goal-directed cognitive process, were substantially effective in decoding the sequential retrieval processes (i.e., backward versus forward retrieval conditions). These findings suggest that the large-scale whole-brain networks are required during sequential retrieval processes of episodic memory, supporting the proposal that controlled processes across cortical and subcortical are recruited for the retrieval of temporal events and are mirrored in the patterns of functional connectivity.
, temporal correlations) of this region with the rest of the brain was assessed in a voxel-wise manner. The results showed a distributed connectivity maps encompassing the brain regions such as the bilateral superior frontal gyri, bilateral middle frontal gyri, bilateral inferior frontal gyri, bilateral medial frontal gyri, bilateral middle/inferior temporal gyri, bilateral caudate, and bilateral inferior parietal lobules (Table 2 and Fig. 4).
To test whether the obtained peak accuracies were statistically different from the null distribution results, we conducted permutation testing which showed that accuracies remained near 50% demonstrating that the peak accuracies from the original data were derived from distinctive patterns of functional connectivity between the two different retrieval conditions (91 feature permutation mean = 0.49, SD = 0.12, C.I (95%) = 0.005; p < .001).
참고문헌 (54)
Anders, T. R., & Lillyquist, T. D. (1971). Retrieval time in forward and backward recall. Psychonomic Science, 22, 205-206.
Badre, D., & D'Esposito, M. (2007). Functional magnetic resonance imaging evidence for a hierarchical organization of the prefrontal cortex. Journal of Cognitive Neuroscience, 19, 2082-2099.
Badre, D., Poldrack, R. A., Pare-Blagoev, E. J., Insler, R. Z., & Wagner, A. D. (2005). Dissociable controlled retrieval and generalized selection mechanisms in ventrolateral prefrontal cortex. Neuron, 47, 907-918.
Barnes, K. A., Cohen, A. L., Power, J. D., Nelson, S. M., Dosenbach, Y. B., Miezin, F. M., Petersen, S. E., & Schlaggar, B. L. (2010). Identifying basal ganglia divisions in individuals using resting-state functional connectivity MRI. Frontiers in Systems Neuroscience, 4, 18.
Barredo, J., Verstynen, T. D., & Badre, D. (2016). Organization of cortico-cortical pathways supporting memory retrieval across subregions of the left ventrolateral prefrontal cortex. J Neurophysiol, 116, 920-937.
Cabeza, R., Ciaramelli, E., Olson, I. R., & Moscovitch, M. (2008). The parietal cortex and episodic memory: an attentional account. Nature Reviews Neuroscience, 9, 613-625.
D'Esposito, M., Postle, B. R., & Rypma, B. (2000). Prefrontal cortical contributions to working memory: evidence from event-related fMRI studies. Experimental Brain Research, 133, 3-11.
Dobbins, I. G., & Han, S. (2006). Cue- versus probe-dependent prefrontal cortex activity during contextual remembering. Journal of Cognitive Neuroscience, 18, 1439-1452.
Friston, K. J., Buechel, C., Fink, G. R., Morris, J., Rolls, E., & Dolan, R. J. (1997). Psychophysiological and modulatory interactions in neuroimaging. Neuroimage, 6, 218-229.
Friston, K. J., Glaser, D. E., Henson, R. N., Kiebel, S., Phillips, C., & Ashburner, J. (2002). Classical and Bayesian inference in neuroimaging: applications. Neuroimage, 16, 484-512.
Han, S., O'Connor, A. R., Eslick, A. N., & Dobbins, I. G. (2012). The role of left ventrolateral prefrontal cortex during episodic decisions: semantic elaboration or resolution of episodic interference? Journal of Cognitive Neuroscience, 24, 223-234.
Koechlin, E., Basso, G., Pietrini, P., Panzer, S., & Grafman, J. (1999). The role of the anterior prefrontal cortex in human cognition. Nature, 399, 148-151.
Kruschwitz, J. D., List, D., Waller, L., Rubinov, M., & Walter, H. (2015). GraphVar: a user-friendly toolbox for comprehensive graph analyses of functional brain connectivity. J Neurosci Methods, 245, 107-115.
Lezak, M. D. (1995). Neuropsychological assessment. New York: Oxford University Press.
Marklund, P., Larsson, A., Elgh, E., Linder, J., Riklund, K. A., Forsgren, L., & Nyberg, L. (2009). Temporal dynamics of basal ganglia under-recruitment in Parkinson's disease: transient caudate abnormalities during updating of working memory. Brain, 132, 336-346.
Moss, H. E., Abdallah, S., Fletcher, P., Bright, P., Pilgrim, L., Acres, K., & Tyler, L. K. (2005). Selecting among competing alternatives: selection and retrieval in the left inferior frontal gyrus. Cerebral Cortex, 15, 1723-1735.
Neubert, F. X., Mars, R. B., Thomas, A. G., Sallet, J., & Rushworth, M. F. (2014). Comparison of human ventral frontal cortex areas for cognitive control and language with areas in monkey frontal cortex. Neuron, 81, 700-713.
Pantazatos, S. P., Talati, A., Schneier, F. R., & Hirsch, J. (2014). Reduced anterior temporal and hippocampal functional connectivity during face processing discriminates individuals with social anxiety disorder from healthy controls and panic disorder, and increases following treatment. Neuropsychopharmacology, 39, 425-434.
Pirogovsky, E., Goldstein, J., Peavy, G., Jacobson, M. W., Corey-Bloom, J., & Gilbert, P. E. (2009). Temporal order memory deficits prior to clinical diagnosis in Huntington's disease. Journal of the International Neuropsychological Society, 15, 662-670.
Raposo, A., Han, S., & Dobbins, I. G. (2009). Ventrolateral prefrontal cortex and self-initiated semantic elaboration during memory retrieval. Neuropsychologia, 47, 2261-2271.
Sagar, H. J., Sullivan, E. V., Gabrieli, J. D., Corkin, S., & Growdon, J. H. (1988). Temporal ordering and short-term memory deficits in Parkinson's disease. Brain, 111, 525-539.
Seeley, W. W., Menon, V., Schatzberg, A. F., Keller, J., Glover, G. H., Kenna, H., Reiss, A. L., & Greicius, M. D. (2007). Dissociable intrinsic connectivity networks for salience processing and executive control. Journal of Neuroscience, 27, 2349-2356.
Shirer, W. R., Ryali, S., Rykhlevskaia, E., Menon, V., & Greicius, M. D. (2012). Decoding subject-driven cognitive states with whole-brain connectivity patterns. Cerebral Cortex, 22, 158-165.
Song, X. W., Dong, Z. Y., Long, X. Y., Li, S. F., Zuo, X. N., Zhu, C. Z., He, Y., Yan, C. G., & Zang, Y. F. (2011). REST: a toolkit for resting-state functional magnetic resonance imaging data processing. PLoS One, 6, e25031.
Spreng, R. N., Stevens, W. D., Chamberlain, J. P., Gilmore, A. W., & Schacter, D. L. (2010). Default network activity, coupled with the frontoparietal control network, supports goal-directed cognition. Neuroimage, 53, 303-317.
Takahashi, E., Ohki, K., & Kim, D. S. (2007). Diffusion tensor studies dissociated two fronto-temporal pathways in the human memory system. Neuroimage, 34, 827-838.
Thomas, J. G., Milner, H. R., & Haberlandt, K. F. (2003). Forward and backward recall: different response time patterns, same retrieval order. Psychological Science, 14, 169-174.
Thompson-Schill, S. L., D'Esposito, M., Aguirre, G. K., & Farah, M. J. (1997). Role of left inferior prefrontal cortex in retrieval of semantic knowledge: a reevaluation. Proceedings of the National Academy of Sciences, U.S.A., 94, 14792-14797.
Thompson-Schill, S. L., D'Esposito, M., & Kan, I. P. (1999). Effects of repetition and competition on activity in left prefrontal cortex during word generation. Neuron, 23, 513-522.
Vincent, J. L., Kahn, I., Snyder, A. Z., Raichle, M. E., & Buckner, R. L. (2008). Evidence for a frontoparietal control system revealed by intrinsic functional connectivity. Journal of Neurophysiology, 100, 3328-3342.
Vriezen, E. R., & Moscovitch, M. (1990). Memory for temporal order and conditional associative-learning in patients with Parkinson's disease. Neuropsychologia, 28, 1283-1293.
Wagner, A. D., Pare-Blagoev, E. J., Clark, J., & Poldrack, R. A. (2001). Recovering meaning: left prefrontal cortex guides controlled semantic retrieval. Neuron, 31, 329-338.
Wechsler, D. (2008). Wechsler adult intelligence scale-fourth edition (WAIS-IV). San Antonio: NCS Pearson.
Yan, C. G., & Zang, Y. F. (2010). DPARSF: a MATLAB toolbox for "pipeline" data analysis of resting-state fMRI. Frontiers in Systems Neuroscience, 4, 13.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.