Table 1.
Overview of objective rehabilitation assessment.
| Paper | Tools | Limb | Method | Assessment objectives |
|---|---|---|---|---|
| Eichler et al. (2018) | 3D cameras, FMA | Whole body | SVM1, SDT2 and RF3 | Distinguish between the three known stroke severity. |
| Wang et al. (2022d) | sEMG, Brunnstrom | Forearms | Ensemble Learning | Brunnstrom stage automatic evaluation for stroke survivor. |
| Postolache et al. (2015) | Doppler radars, Gait sensors | Legs, Feet | Wavelet multiresolution analysis | Assess the instantaneous velocity of the leg swing. |
| Park et al. (2020) | IMU, NIHSS and MRC | Wrists, Ankles and Feet | Ensemble algorithm, SVM | Automatically assess stroke survivor' NIHSS grades and MRC scores. |
| Zhao et al. (2017) | IMU, Pressure sensors | Knees, Ankles and Feet | Inequality-constrained Zero Velocity Updates-aided Inertial Navigation System algorithm | Foot Angle for Assessing Extension-Flexion Movement in Stroke survivors. |
| Guo et al. (2019) | sEMG, A State-Space EMG Model (Han et al., 2015) | Upper limbs | Bayesian classifier | Distinguishing stages of stroke rehabilitation. |
| Zeng et al. (2017) | EEG | Corresponding brain area of the lesion | Mean-NLSD4, PSD5 | Assess stroke recovery. |
| Wang et al. (2022b) | sEMG, EEG | Corresponding brain area of the lesion, Dorsal interosseous | PTE6 | Using brain network features and cortical muscle coupling values to distinguish rehabilitation stages. |
| Bervet et al. (2013) | EMG, Cameras and Pressure sensors | Feet, Legs and Heels | KeR-EGI7 | Gait recognition on healthy subjects. |
| Meng et al. (2021) | sEMG, IMU | Weist, Forearms, Legs and Ankles | LOSOCV8 | Mobility capacity in a stroke survivor. |
| Spanos et al. (2023) | Cameras, G.A.I.T | Hip, Legs and Ankles | KNN9 | Gait recognition on healthy subjects. |
| Wang et al. (2020) | Cameras, IMU | Legs and Ankles, Feet | INI10 | judge the recovery of each leg. |
1
SVM, Support Vector Machines;
2
SDT, Single Decision Tree;
3
RF, Random Forest;
4
Mean-NLSD, Mean-Nonlinearly Separable Degree;
5
PSD, power Spectral Density;
6
PTE, Phase Transfer Entropy;
7
KeR-EGI, Kerpape-Rennes EMG-based Gait Index;
8
LOSOCV, Leave-One-Subject-Out Cross-Validation;
9
KNN, K Nearest Neighbor;
10
INI, IMU-based gait normalcy index.