PDF(Pubmed)
Abstract：
We tested the hypothesis, generated from the Gradient Order Directions Into Velocities of Articulators (GODIVA) model, that adults who stutter (AWS) may comprise subtypes based on differing connectivity within the cortico-basal ganglia planning or motor loop. Resting state functional connectivity from 91 AWS and 79 controls was measured for all GODIVA model connections. Based on a principal components analysis, two connections accounted for most of the connectivity variability in AWS: left thalamus - left posterior inferior frontal sulcus (planning loop component) and left supplementary motor area - left ventral premotor cortex (motor loop component). A k-means clustering algorithm using the two connections revealed three clusters of AWS. Cluster 1 was significantly different from controls in both connections; Cluster 2 was significantly different in only the planning loop; and Cluster 3 was significantly different in only the motor loop. These findings suggest the presence of planning and motor subtypes of stuttering.
摘要：
我们检验了假设,从关节器速度的梯度顺序方向(GODIVA)模型生成，口吃的成年人（AWS）可能包含基于皮质-基底神经节计划或运动回路内不同连通性的亚型。对于所有GODIVA模型连接，测量了来自91个AWS和79个对照的静息状态功能连接。基于主成分分析，两个连接占AWS中连接变异性的大部分：左丘脑-左后下额沟（计划环路组件）和左补充运动区-左腹前运动皮层（运动环路组件）。使用两个连接的k均值聚类算法揭示了AWS的三个集群。群集1在两个连接中都与控件显着不同；群集2仅在计划循环中存在显着差异；群集3仅在运动循环中存在显着差异。这些发现表明存在计划和口吃的运动亚型。
