Skip to main content

Table 3 Correlations across gait quality measures

From: Hidden Markov model-based similarity measure (HMM-SM) for gait quality assessment of lower-limb prosthetic users using inertial sensor signals

 

GPS

HMM-SM upper

HMM-SM lower

HMM-SM pelvis

MDP upper

MDP lower

MDP pelvis

DTW upper

DTW lower

DTW pelvis

INI

HMM-SM upper

0.428 (0.06)

          

HMM-SM lower

−0.549* (0.011)

−0.294 (0.213)

         

HMM-SM pelvis

−0.350 (0.138)

−0.339 (0.153)

0.412 (0.07)

        

MDP upper

0.290 (0.214)

−0.031 (0.908)

−0.167 (0.471)

−0.197 (0.413)

       

MDP lower

0.491* (0.028)

0.188 (0.43)

−0.411 (0.07)

−0.285 (0.218)

0.860* (<0.001)

      

MDP pelvis

0.567* (0.008)

0.145 (0.52)

−0.470* (0.036)

−0.319 (0.181)

0.720* (<0.001)

0.748* (<0.001)

     

DTW upper

0.328 (0.168)

−0.018 (0.944)

−0.219 (0.358)

−0.147 (0.52)

0.947* (<0.001)

0.815* (<0.001)

0.757* (<0.001)

    

DTW lower

0.677* (<0.001)

0.147 (0.52)

−0.535* (0.014)

−0.256 (0.278)

0.594* (0.005)

0.832* (<0.001)

0.730* (<0.001)

0.611* (0.003)

   

DTW pelvis

0.575* (0.007)

0.167 (0.471)

−0.518* (0.018)

−0.239 (0.316)

0.677* (<0.001)

0.705* (<0.001)

0.968* (<0.001)

0.733* (<0.001)

0.718* (<0.001)

  

INI

0.761* (<0.001)

0.196 (0.413)

−0.415 (0.07)

−0.296 (0.213)

0.234 (0.323)

0.392 (0.087)

0.443 (0.052)

0.295 (0.213)

0.645* (0.002)

0.432 (0.059)

 

MGS

−0.487 (0.029)

−0.185 (0.431)

0.533* (0.014)

0.007 (0.972)

−0.039 (0.889)

−0.389 (0.088)

−0.286 (0.218)

−0.097 (0.679)

−0.578* (0.007)

−0.306 (0.202)

−0.449* (0.049)

  1. R-values listed for each comparison, with corresponding adjusted p-values in parentheses. Correlations with achieved significance threshold (alpha = 0.05) are bolded and have an asterisk (*). The first column shows results comparing the inertial sensor-based measures to the GPS, while the rest of the cells indicate comparisons between the different inertial sensor-based measures. Of note, the MGS-GPS comparison is not highlighted, as the correlation was significant but in the incorrect direction as expected
  2. GPS Gait Profile Score, HMM-SM hidden Markov model-based similarity measure, MDP Movement deviation profile, DTW Dynamic time warping, INI IMU-based Gait Normalcy Index, MGS Multifeature Gait Score