科研笔记 -- HDsEMG数据集分析(1)

最近看到一篇做的很好的高密度肌电信号数据集,简单分析以下。
名字是:Muscles from the same muscle group do not necessarily share common drive

采集部位

采集了三块肌肉:

  1. 腓肠肌内侧(GM) – 64电极
  2. 腓肠肌外侧(GL) – 64电极
  3. 比目鱼肌(SOL)的内侧部分 – 32电极

figure 1

在这里插入图片描述

原文描述
Figure 1. Experimental setup. A: shows the placement of the high-density electromyography electrodes. A two-dimensional adhesive grid of 64 electrodes was placed over both the gastrocnemius medialis (GM) and the gas-trocnemius lateralis (GL) muscle. A two-dimensional adhesive grid of 32 electrodes was placed on the medial part of the soleus (SOL), below the myotendinous junction of the GM muscle. B: shows the isometric plantar-flexion task performed on the ergometer (knee was fully extended and ankle angle at 10 of plantarflexion). C: shows the isometric heel-raise task which consisted of holding the heel at a standard height of 6cm, with either the feet in neutral position (comfortable stance; toes neutral) or internally rotated (toes in).
图1.实验装置。A:显示高密度肌电图电极的位置。将64个电极的二维粘合网格放置在腓肠肌内侧(GM)和腓肠肌外侧(GL)肌肉上。将32个电极的二维粘合网格放置在比目鱼肌(SOL)的内侧部分,位于GM肌肉的肌腱连接处下方。B:显示在测力计上进行的等长跖屈任务(膝关节完全伸展,踝关节角度为跖屈10 °)。丙:示出了等长的脚跟抬高任务,其包括将脚跟保持在6cm的标准高度,脚处于中立位置(舒适的站立;脚趾中立)或内旋(脚趾向内)。

文章分析

先把该论文中剩余的图表分析一下:

figure 2 平滑&瞬时放电速率

![在这里插入图片描述](https://img-blog.csdnimg.cn/direct/a6a1421c5cd84b7b851d17133ffc41f1.png

原文描述:
Figure 2. Individual example of the assessment of the proportion of common input (PCI). A: shows the smoothed discharge rate (cut-off frequency =2Hz) of 25 motor units identified from the gastrocnemius medialis, during the isometric heel-raise task performed with the toes in neutral position, for participant 12.Note that only 20s of the contraction are depicted here. B: shows the results of the coherence analyses performed on two cumulative spike trains, with varying number of motor units in each group, i.e., from 1 to 12 motor units. Each estimation is the average of 100 random permutations of motor units. The horizontal line indicates the threshold of significant coherence for 12 motor units per group. C:represents the relationship between the mean values (±standard deviation) of coherence in the bandwidth 0–5Hz and the number of motor units. This relationship was fitted by least-squares (red line) and the rate of change was considered as an index of the proportion of common input (PCI).
实例评估了个体共同投入比例(PCI)。A:显示了参与者12在脚趾处于中性位置的情况下进行等长脚跟抬高任务期间,从腓肠肌内侧识别的25个运动单元的平滑放电速率(截止频率= 2 Hz)。注意,这里仅描绘了20 s的收缩。B:示出了对两个累积尖峰序列执行的相干性分析的结果,其中每组中具有不同数量的运动单元,即,从1到12个运动单位。每个估计是运动单位的100个随机排列的平均值。水平线表示每组12个运动单元的显著相干性的阈值。C:表示0- 5 Hz带宽内相干性的平均值(±标准差)与运动单位数之间的关系。用最小二乘法(红线)拟合该关系,并将变化率视为共同输入比例(PCI)的指标。

figure 3

A:
B:显示GM的一个MU的瞬时放电率估计的三个结果
在这里插入图片描述

原文描述
Figure 3. Individual example of the assessment of the neural drive during the isometric plantarflexions performed on the ergometer. A: represents the smoothed discharge rate (cut-off frequency=2Hz) of the motor units identified in participant 11 during the contraction at 30% of MVC. B:
shows three outcomes (time at recruitment, d, and peak discharge rate) estimated from the instantaneous discharge rate of one motor unit of the gastrocnemius medialis muscle. Note that the peak discharge rate
value was estimated from a sixth-order polynomial used to smooth the instantaneous discharge rates. C: shows the cumulative spike train (CST; index of neural drive) of each of the threemuscles. GL, gastrocnemius lateralis; GM, gastrocnemius medialis; MVC, maximal voluntary contraction; SOL, soleus (medial-posterior compartment).
图3.在测力计上进行等长跖屈时神经驱动评估的个体示例。A:表示在30%MVC的收缩期间在参与者11中识别的运动单位的平滑放电速率(截止频率= 2 Hz)。B:显示了根据腓肠肌内侧肌的一个运动单位的瞬时放电率估计的三个结果(募集时间、d和峰值放电率)。注意,峰值放电率值是根据用于平滑瞬时放电率的六阶多项式估计的。C:显示三块肌肉中每一块的累积尖峰序列(CST;神经驱动指数)。GL,腓肠肌外侧; GM,腓肠肌内侧; MVC,最大自主收缩; SOL,比目鱼肌(内-后室)。

Table 1

在这里插入图片描述

原文描述
Because of technical reasons, there were no data for participant 14 during the isometric plantarflexions and for participant 2 during the heel raise with the toes in neutral condition. In addition, soleus recordings are missing for one and two participants during the plantarflexions and the heel-raise conditions, respectively. Consequently, GM and GL data are reported for n=17 participants (isometric plantarflexions and isometric heel raise with the toes in neutral condition); SOL data are reported for n=16 (isometric plantarflexions) and n=15 (isometric heel raises) participants. GL, gastrocnemius lateralis; GM, gastrocnemius medialis; MVC, maximal voluntary contractions; SOL, soleus.
由于技术原因,参与者14在等长跖屈期间没有数据,参与者2在脚趾处于中性状态的脚跟抬起期间没有数据。此外,比目鱼肌记录丢失的一个和两个参与者在跖屈和脚跟抬起的条件下,分别。因此,报告了n=17名参与者的GM和GL数据(等长跖屈和等长足跟抬高,脚趾处于中性状态);报告了n=16名(等长跖屈)和n=15名(等长足跟抬高)参与者的SOL数据。GL,腓肠肌外侧; GM,腓肠肌内侧; MVC,最大自主收缩; SOL,比目鱼肌。

figure 4

在这里插入图片描述

原文描述
Proportion of common synaptic input to motor neurons within the same muscle, estimated during the isometric heel raise (toes neutral). The within-muscle coherence analysis was performed on 17, 11, and 10 participants for gastrocnemius medialis (GM), gastrocnemius lateralis (GL), and soleus (SOL), respectively. Each participant is represented by a different color and the horizontal line indicates the mean value. ¥, indicates significant difference compared with SOL. There was no significant difference between GM and GL.
相同肌肉内运动神经元的共同突触输入比例,在等长脚跟抬起(脚趾中立)期间估计。分别对17、11和10名参与者的腓肠肌内侧肌(GM)、腓肠肌外侧肌(GL)和比目鱼肌(SOL)进行肌肉内连贯性分析。每个参与者用不同的颜色表示,水平线表示平均值。$,表示与SOL相比有显著差异。GM和GL之间无显著差异。

figure 5

在这里插入图片描述

原文描述
Figure 5. Between-muscle coherence estimated during the isometric heel raise (toes neutral). A: shows examples of the coherence between GM and GL for two participants: participant 18 who exhibited no coherence and participant 5 who exhibited the highest coherence over the population tested. The horizontal lines represent the threshold of significance defined as the highest coherence value for frequencies>100Hz, at which no coherence is expected. The other panels show, for each participant, the frequencies at which a significant coherence was observed for GM-GL (B), for GM-SOL ©, and for GL-SOL (D). Some participants were discarded from this analysis as too few motor units were discriminated to allow this analysis. Individual coherence data are depicted in Figs. 6. GL, gastro- cnemius lateralis; GM, gastrocnemius medialis;SOL,soleus.
图5.在等长脚跟抬起(脚趾中立)期间估计的肌肉间连贯性。A:显示了两个参与者的GM和GL之间的一致性的例子:参与者18没有表现出一致性,参与者5在测试的人群中表现出最高的一致性。水平线表示被定义为频率> 100 Hz的最高相干性值的显著性阈值,在该频率下不期望相干性。其他图显示了每个参与者观察到GM-GL(B)、GM-SOL(C)和GL-SOL(D)显著一致性的频率。一些参与者被从该分析中排除,因为太少的运动单位被区分出来,无法进行该分析。个体相干性数据描绘于图1A和1B中。6.GL,腓肠肌外侧; GM,腓肠肌内侧;SOL,比目鱼肌。

figure 6

在这里插入图片描述

原文描述
Figure 6. Individual data of coherence between the three heads of the triceps surae estimated during the isometric heel raise (toes neutral). The red horizontal lines represent the threshold of significance defined as the highest coherence value for frequencies>100Hz, at which no coherence is expected. Some participants were discarded from this analysis as too few motor units were discriminated to allow this analysis. GL, gastrocnemius lateralis; GM, gastrocnemius medialis; SOL, soleus. A: GM-GL coherence; B: GM-SOL coherence; C:GL-SOL coherence.
图6.在等长足跟抬高(脚趾中立)时估计的小腿三头肌三头之间连贯性的个体数据。红色水平线表示被定义为频率>100赫兹的最高相干值的重要阈值,在该值时不需要相干性。一些参与者被排除在这项分析之外,因为被区别对待的运动单位太少,不能进行这项分析。GL,腓肠肌外侧肌;GM,腓肠肌内侧肌;SOL,比目鱼肌。A:GM-GL一致性;B:GM-SOL一致性;C:GL-SOL一致性。

figure 7

在这里插入图片描述

原文描述
Figure 7. Motor unit discharge characteristics during the isometric plantarflexion tasks performed at 30% (A) and 50% ofMVC (B). Each participant is represented by a different color, and the horizontal bar indicates the mean value. The dashed horizontal line on the upper panels indicates the beginning of the torque plateau. P< 0.05 for comparison with GM; $ P< 0.05 for comparison with SOL; £P< 0.05 for comparison with GL. GL, gastrocnemius lateralis; GM, gastrocnemius medialis; MVC, maximal voluntary contractions; SOL, soleus.
图7.在30% MVC(A)和50% MVC(B)条件下进行等长跖屈任务时的运动单位放电特征。每个参与者用不同的颜色表示,水平条表示平均值。上图上的水平虚线表示扭矩平稳期的开始。与GM比较P< 0.05;与SOL比较$P< 0.05;与GL比较P< 0.05。GL,腓肠肌外侧; GM,腓肠肌内侧; MVC,最大自主收缩; SOL,比目鱼肌。

figure 8 放电速率

在这里插入图片描述

原文描述
Figure 8. Discharge rate during the isometric heel raise performed with different feet configurations. Each participant (n=14–17) is represented by a different color, and the horizontal bar indicates the mean value. P < 0.05 for comparison with toes neutral. A: gastrocnemius medialis; B: gastrocnemius lateralis.
图8.在不同的脚配置进行等长脚跟抬起过程中的放电速率。每个参与者(n=14-17)用不同的颜色表示,水平条表示平均值。P < 0.05,与脚趾中性比较。A:内侧腓肠肌; B:外侧腓肠肌。

R语言实战笔记第九章介绍了方差分析的内容。方差分析是一种用于比较两个或多个组之间差异的统计方法。在R语言中,可以使用lm函数进行方差分析的回归拟合。lm函数的基本用法是: myfit <- lm(I(Y^(a))~x I(x^2) I(log(x)) var ... [-1],data=dataframe 其中,Y代表因变量,x代表自变量,a代表指数,var代表其他可能对模型有影响的变量。lm函数可以拟合回归模型并提供相关分析结果。 在方差分析中,还需要进行数据诊断,以确保模型的可靠性。其中几个重要的诊断包括异常观测值、离群点和高杠杆值点。异常观测值对于回归分析来说非常重要,可以通过Q-Q图和outlierTest函数来检测。离群点在Q-Q图中表示落在置信区间之外的点,需要删除后重新拟合并再次进行显著性检验。高杠杆值点是指在自变量因子空间中的离群点,可以通过帽子统计量来识别。一般来说,帽子统计量高于均值的2到3倍即可标记为高杠杆值点。 此外,方差分析还需要关注正态性。可以使用car包的qqplot函数绘制Q-Q图,并通过线的位置来判断数据是否服从正态分布。落在置信区间内为优,落在置信区间之外为异常点,需要进行处理。还可以通过绘制学生化残差的直方图和密度图来评估正态性。 综上所述,R语言实战第九章介绍了方差分析及其相关的数据诊断方法,包括异常观测值、离群点、高杠杆值点和正态性检验。这些方法可以用于分析数据的可靠性和模型的适应性。<span class="em">1</span><span class="em">2</span><span class="em">3</span> #### 引用[.reference_title] - *1* *2* *3* [R语言实战笔记--第八章 OLS回归分析](https://blog.csdn.net/gdyflxw/article/details/53870535)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_2"}}] [.reference_item style="max-width: 100%"] [ .reference_list ]
评论 1
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值