| 摘要: | 在我們的日常生活中，不管是工作還是學習，工作記憶都起到了十分重要的作用，因此工作記憶能力的評估和訓練也隨之成為學者研究的重點。目前來說，對於工作記憶上的評估，一般都採用智力測驗，但是行為資料（智力測驗等）真的可以完全評估當時受試者的神經生理變化？工作記憶訓練真的可以改變我們的大腦嗎？研究首先利用n-back試驗時ERP成份N200和P300與其成績比較，尋找腦波與外在行為的關聯性，並且用N200去評估工作記憶訓練的成效。
本研究招募了三十位中央大學學生為受試者，視力正常且無中樞神經疾病病史，慣用手均為右手。實驗組、對照組（1）與對照組（2）分別有十人。實驗組被要求進行一個與spatial span作業相似的益智遊戲，為期三周，每週五日，每日三十分鐘。對照組（1）則進行與記憶無關的動作控制遊戲，訓練頻率與實驗組同。而對照組（2）不需要接受任何訓練。再由3篇文獻為基礎，建立3種不同的模型，去尋找最適合本研究的神經網路結構。最後，由最佳網路結構去發現訓練前後哪些區域連接出現顯著變化。
用事件相關電位與成績來尋找工作記憶較強的腦波，然而並沒有找到有利的依據。以網頁益智遊戲作為基礎的記憶訓練，從事件相關電位所得到的電生理資料上，3個電極位（C4、T8、P8）出現實驗組N2振幅的增加數值顯著高於對照組，這也表明記憶訓練確有其效果，而且其效果至少可維持一周。小魚急急棒的訓練，並不能對工作記憶產生顯著的影響。
經過DCM的處理，我們發現以網頁益智遊戲作為基礎的記憶訓練，使得我們更合理的去分配大腦的資源。記憶訓練不僅使得Forward中LIPS連接SFG、SFG連接RMFG，Backward中RIPS連接RV1，Lateral中LIPS連接RIPS的連接強度得到顯著增加，還抑制了一些我們不太需要用到的腦區連接，如RMFG連接SFG。
;In our daily lives, working memory (WM) has played a very important role during working or learning. So many studies have focused on WM capability assessment and training of WM. For now, researchers generally use the intelligence tests to assess the WM capacity. But, can the behavioral data (Intelligence tests etc.) completely assess the neurophysiological change? Does the working memory training really change our brains? To answer the questions, we first correlated the performance of the N-back task to its ERPs (P300 and N200) to find their relation and then the amplitude and latency of N200 to assess the effectiveness of working memory training. In addition, DCM was employed to explore the mechanisms underlying the n-back task before and after WM training.
This study recruited thirty Central University student subjects with normal vision and no history of central nervous system disorders, and they are right handedness. Training group, control group (1) and control group (2) respectively have 10 persons. Subjects in the training group were asked to perform spatial span in three weeks, five days a week and thirty minutes a day. Subjects in the control group (1) need to complete motion control game, it is independent of memory, with the experimental group trained the same frequency. While the control group (2) does not require any training accepted. And then by the three documents as the basis, we have established three different models, to find the most suitable for the study of the neural network structure. Finally, the best structure to find out which areas of the network connection appears significant change before and after the training.
There is no significant relation between ERPs and the performance of the n-back task. Memory training of web puzzle game, we find three electrode sites (C4, T8, P8) appear the value of N2 amplitudes increased in the experimental group are significantly higher than control group, This shows that memory training has its positive effects, it can indeed reduce the impact of memory load on the brain, and its effects can be maintained for at least a week. The fish sticks hastily trained cannot produce a significant impact on the working memory.
Memory training not only makes the connection strength to get a significant increase in some brain regions , for example LIPS connects SFG and SFG connects RMFG in forward, RIPS connects RV1 in backward, LIPS connects RIPS in lateral, also inhibits some we do not need to use the brain area connection, such as RMFG connection SFG. |
