大腦及小腦皮質在健康成人與非小腦損傷缺血性中風病患之短期與長期腳踝動作次序學習扮演的角色: 功能性磁振造影研究

Roles of the Cerebral and Cerebellar Cortices in Short- and Long-term Ankle Motor Sequence Learning of Healthy Adults and Patients with Non-cerebellar Ischemic Stroke: An fMRI Study

陳亦璇(Yi-Hsuan Chen)

指導教授 ： 湯佩芳

國立臺灣大學/醫學院/物理治療學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201700421

摘要

背景：中風後之動作學習以及其相關腦部神經塑性為現今腦中風復建之重要議題。過去針對次序性動作學習及其相關神經塑性之研究乃建立在使用上肢進行短期動作學習的模式上。然而，不同的大腦及小腦功能區在短期與長期腳踝動作次序性動作學習的角色仍未被完整探討。
目的：本研究之目的為探討非小腦損傷之缺血性中風病人以及年齡相當之健康成年人之大腦及小腦皮質在短期（一週）及長期（三週）之次序性動作學習後的活化改變，以及腦功能區活化改變和行為改變上之相關性。
方法：本研究使用三週之腳踝動作次序學習，收入非小腦損傷之慢性缺血性中風病人及年齡相當之健康成年人為受試者，使用非慣用腳或患側腳腳踝進行學習。於開始訓練前、開始訓練一週後及開始訓練三週後分別執行行為測試以及功能性磁振造影之掃描。原始之功能性磁振造影影像先以前處理(preprocessing)之步驟去除雜訊。接著會進行全腦分析(whole brain analysis)，利用對比之設定，分析每位受試者執行重複次序時的腦部功能性活化與休息時的腦部功能性活化差異。本研究採用依解剖自動標籤模板（Anatomical Automatic Labeling template）選取之雙側前額葉、感覺動作皮質區、前運動皮質區、輔助動作皮質區、小腦小葉IV&V、VI、VII、VIII及Crus I+II等為興趣區域（regions of interest）。探討不同的興趣區在學習前、學習一週後、及學習三週後之功能性活化模式的改變，及學習後興趣區域腦部活化改變以及行為改變間的相關性。
結果：本研究納入20位健康人以及18位中風受試者進入短期學習組，12位健康人以及10位中風受試者進入長期學習組。腳踝追蹤結果顯示，經過一週及三週之學習過後，兩組受試者均在行為表現有統計上顯著之進步(p< 0.05)。影像結果顯示，健康受試者在學習前主要使用兩側之輔助動作皮質區、前運動皮質區、主要感覺動作皮質區、前額葉及小腦小葉VI執行腳踝重複次序動作(p< 0.05)。中風受試者則比健康受試者多使用動作對側小腦小葉IV&V (p< 0.05)。一週學習過後，健康受試者在動作同側大腦輔助動作皮質區、前運動皮質區、動作對側主要感覺動作皮質區與動作同側前額葉（r= 0.486~0.534, p< 0.05）及額外兩週學習過後，於大腦前動作對側運動皮質區、主要感覺動作皮質區、雙側小腦小葉IV&V、動作對側小腦小葉VI、小腦小葉Crus I&II（r= 0.404~0.453, p> 0.05）之標準化活化強度下降越多者，標準化行為進步越多。一週學習過後，中風受試者在小腦小葉動作同側VII及VIII之標準化活化強度下降越少者，行為表現進步越多（r= -0.643~-0.684, p< 0.05）。經過額外兩週的練習，中風受試者在動作對側大腦主要感覺動作皮質區、雙側前運動皮質區、雙側前額葉（r= 0.652~0.789, p> 0.05）之標準化活化強度下降越多者及小腦小葉動作對側VI,雙側VII及動作對側Crus I&II（r= -0.601~-0.803, p> 0.05）之標準化活化強度上升越多者，標準化行為進步越多。
討論及結論：本研究之結果顯示，大腦與小腦皮質在健康人及非小腦損傷中風病人之短期及長期腳踝動作次序性學習中，扮演不同之角色。健康人短期學習主要倚賴大腦活化效率之提升，其長期學習則倚賴大腦及小腦活化效率之提升。而非小腦損傷中風病人之短期及長期學習則皆主要倚賴小腦之代償性活化。

關鍵字

腦中風 ； 下肢 ； 動作次序性學習 ； 功能性磁振造影 ； 大腦及小腦活化模式

並列摘要

Background: Motor learning and the associated brain plasticity after stroke are crucial topics in stroke rehabilitation nowadays. However, the evidence of the benefits of motor sequence learning was established mainly on studies using upper extremity short-term motor learning paradigms. In previous studies, the roles of the different regions in the cerebrum and the cerebellum in short- and long-term ankle motor sequence learning have not been fully explored.
Purposes: The purposes of this study were to investigate the changes of cerebral and cerebellar functional activation and their relationships with behavioral improvement after short- and long-term ankle motor sequence learning in patients with non-cerebellar ischemic stroke and age-matched healthy adults.
Methods: A three-week motor sequence learning paradigm using the non-dominant or paretic ankle was used in this study. Patients with non-cerebellar ischemic stroke and age-matched healthy controls were enrolled and underwent clinical assessments, behavior testing, functional magnetic resonance imaging (fMRI) scans at baseline, Week 1, and Week 3. Raw fMRI data was preprocessed to eliminate the noise. Then, whole brain analysis (contrast of repeated sequences versus rest condition) and anatomical regions of interest (ROIs) analysis were conducted. Bilateral dorsolateral prefrontal cortex (dlPFC), sensorimotor cortex (SMC), supplementary motor area (SMA), and premotor cortex (PMC), cerebellar lobules IV&V, VI, VII, VIII, and Crus I&II, based on the Anatomical Automatic Labeling (AAL) template, were chosen as the ROIs.
Results: Twenty healthy and 18 stroke subjects participated in the short-term learning study; and 12 healthy and 10 stroke subjects participated in the long-term learning study. Both groups showed significant improvement in ankle tracking errors in short- and long-term learning (p< 0.05). At baseline, healthy subjects showed primary activations in the bilateral SMA, PMC, SMC, contralateral dlPFC, and bilateral cerebellar lobules VI (p< 0.05); the stroke subjects showed additional activations in the contralateral cerebellar lobule IV&V at baseline (p< 0.05). Healthy subjects who had greater reduction of activation intensity in ipsilateral dlPFC, SMA, contralateral SMC, and bilateral PMC from baseline to Week 1 (r= 0.486~0.534, p< 0.05) and in contralateral SMC, PMC, cerebellum lobule VI, lobule Crus I&II, and bilateral cerebellum lobule IV&V from Week 1 to Week 3 (r= 0.404~0.453, p> 0.05) presented greater tracking performance improvement during the corresponding time periods. Stroke patients who had greater increases in activation intensity in ipsilateral cerebellar lobule VII and lobule VIII presented more behavior improvement from baseline to Week 1 ( r= -0.643~-0.684, p< 0.01) and those who showed greater reduction of activation intensity in contralateral SMC, bilateral dlPFC and PMC that from Week 1 to Week 3 (r= 0.652~0.789, p> 0.05) or greater increases in activation intensity in contralateral cerebellum lobule VI, lobule Crus I&II, and bilateral cerebellum lobule VII (r= -0.601~-0.803, p> 0.05) presented more behavior improvement.
Discussion and Conclusions: These results suggested differential roles of the cerebral and cerebellar cortices in short- and long-term ankle motor sequence learning between healthy subjects and patients with chronic stroke with the cerebellum spared. Healthy subjects relied on the efficiency of cerebral activation for short-term learning and on that of cerebral and cerebellar activation for long-term learning. Stroke patients relied on compensatory cerebellar activation for both short- and long-term learning.

並列關鍵字

stroke ； lower extremities ； motor sequence learning ； functional MRI ； cerebral and cerebellar activation pattern

參考文獻

Ashburner J, Friston KJ (2005): Unified segmentation. Neuroimage 26:839-51.

Ashe J, Lungu OV, Basford AT, Lu X (2006): Cortical control of motor sequences. Curr Opin Neurobiol 16:213-21.

Berghuis KM, Veldman MP, Solnik S, Koch G, Zijdewind I, Hortobagyi T (2015): Neuronal mechanisms of motor learning and motor memory consolidation in healthy old adults. Age (Dordr) 37:9779.

Bernard JA, Seidler RD (2013): Cerebellar contributions to visuomotor adaptation and motor sequence learning: An ALE meta-analysis. Frontiers in Human Neuroscience 7:7-27.

Bischoff-Grethe A, Goedert KM, Willingham DT, Grafton ST (2004): Neural substrates of response-based sequence learning using fMRI. J Cogn Neurosci 16:127-38.