以自迴歸模型分析神經元訊號間之因果關係

A Study on the Causality between the Signals of Neurons by the Autoregressive Model

邵培強(Pei-chiang Shao)

指導教授 ： 單維彰

國立中央大學/理學院/數學系/碩士(2011年)

摘要

本研究以麻醉狀況下大鼠前扣帶迴腦皮質 (ACC) 及紋狀體 (STR) 兩腦區於注射「多巴胺第 II 與第 III 類受體高專一性的作用劑」(quinpirole) 的實驗數據為例，利用時間序列的因果分析對神經元之間的連結關係作推論。
分析結果顯示，(1) ACC 內部神經元之間於注射藥物前已有連結，且此連結於注射藥物後穩定地增強。(2) STR 內部神經元之間於注射藥物前後皆無顯著因果關係。(3) ACC 與 STR 之間的神經元於注射藥物前無顯著因果關係，但在注射藥物後產生了由 ACC 至 STR 的資訊流。由於此資訊流的強度較低，我們推測 ACC 與 STR 神經元之連結是間接的，可能透過另一腦核區的神經元。
最後我們依照資訊傳遞的順序將神經元分為「源頭」、「中間」與「目標」三類，並且於注射藥物前後各建構一個神經元子網絡圖。

關鍵字

自迴歸模型 ； 神經元訊號 ； 因果性指標

並列摘要

In this study, we take the neural data which recorded signals from anterior cingulate cortex (ACC) and striatum (STR) during the period of acute quinpirole administration in anesthetized rats as an example, to make inferences on the connections between neurons by determining causality in time series.
The results show that, (1) The neurons in ACC were linked before quinpirole administration, and there was a steady increase in the intensity of these linkage after quinpirole administration. (2) There were no significant causal relationships between STR neurons before nor after quinpirole administration.(3) There were no significant causal relationships between ACC and STR neurons before quinpirole administration, but information flows from ACC to STR arose after quinpirole administration.
Because of the low intensity of these information flows, we conjecture that the connections between ACC and STR neurons are mediated by some undetected neurons.
Finally, neurons were classified into "source", "middle" and "target" three groups according to the order of information delivery. And we construct individually a neural subnetwork before and after quinpirole administration.

並列關鍵字

signals of neurons ； GCI ； autoregressive model

參考文獻

1. W. Hesse, E. Moller, M. Arnold, and B. Schack, The use of time-variant EEG Granger causality for inspecting directed interdependencies of neural assemblies, Journal of Neuroscience Methods 124 (2003) 27-44.

2. E. Moller, B. Schack, M. Arnold, and H. Witte, Instantaneous multivariate EEG coherence analysis by means of adaptive high-dimensional autoregressive models, Journal of Neuroscience Methods 105 (2001) 143-58.

3. A. J. Cadotte, T. B. DeMarse, P. He, and M. Ding, Causal measures of structure and plasticity in simulated and living neural networks, PLoS ONE 3 (2008).

4. L. A. Baccala and K. Sameshima, Partial directed coherence : a new concept in neural stucture determination, Biological Cybernetics 84 (2001) 463-74.

5. L. A. Baccala and K. Sameshima, Overcoming the limitations of correlation analysis for many simultaneously processed neural structures, Progress in Brain Research 130 (2001).