Table 1.
Summary of the selected study characteristics.
| Study | Subjects | Age (yrs) | CP type & distribution | GMFCS | N strides analyzed | N muscles (= total) | EMG pre-processing |
Analysis criteria |
Synergies | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Total N | VAF1 (%) | Walk-DMC | Structure | |||||||||
| Cappellini et al. (2016) | CP: 35 TD: 33 |
CP: 2.3–11.8 TD: 1.0–11.8 |
Spastic (16 uni, 19 bi) |
I, II, III | CP: 50 ± 24 TD: 72 ± 28 |
11 bi (= 22) | HP: 30 Hz, Demeaned LP: 10 Hz |
RMSE of VAF vs. n curve <10−4 | CP: 4 TD: 4 |
- | - | Temporal: CP ≠ TD |
| Cappellini et al. (2018) | CP: 14 TD: 14 |
CP: 3.0–11.1 TD: 3.3–11.8 |
Spastic (5 uni, 9 bi) |
I, II | CP: 35 ± 5 TD: 27 ± 3 |
11 bi (= 22) |
HP: 30 Hz, LP: 10 Hz |
RMSE of VAF vs. n curve <10−4 |
CP: 4 TD: 4 |
- | - | Temporal: CP ≠ TD |
| Hashiguchi et al. (2018) | CP: 13 TD: 10 |
CP: 12.8 ± 3.8 TD: 13.4 ± 0.5 |
NG | I, II, III | 5 | 8 uni (= 8) |
BP: 20–250 Hz, LP: 10 Hz |
VAF>90% | CP: 55% = 2, 30% = 3, 15% = 4 TD: 10% = 3, 60% = 4, 30% = 5 |
- | - | - |
| Tang et al. (2015) | CP: 12 TD: 8 AD: 10 |
CP: 5.8 (3.7–9.0) TD: 6.1 (4.5–9.2) AD: 24.5 (23–26) |
Spastic (2 uni, 9 bi) 1 Dysk |
I, II, III, IV | >20 | 8 bi (= 16) |
HP: 50 Hz, Demeaned, LP: 10 Hz |
VAF>95% | CP: 37.5% = 2*, 29.2% = 3*, 33.3% = 4* TD: 31.2% = 3, 68.8% = 4 AD: 100% = 4 |
- | - | CP ≠ TD&AD SCA: CP = 57.0 ± 16.8, TD = 84.2 ± 11.8, AD = 95.7 ± 2.0 |
| Yu et al. (2019) | CP: 18 TD: 8 |
CP: 4.4 (2.3–6.5) TD: 4.4 ± 1.4 |
Spastic (bi) | I, II, III | 8 (NMF on each stride separately) | 8 bi (= 16) |
HP: 50 Hz, Demeaned, LP: 10 Hz |
VAF4 | CP: GMFCS I/II = 4, GMFCS III = 3 TD: 4 |
- | - | Spatial: CMFCS I/II = TD GMFCS III ≠ TD Temporal: CP ≠ TD |
| Torricelli et al. (2014) | CP: 3 | 15, 14, 14 | Spastic (bi) | II | >3 | 8 bi (= 16) |
BP: 20–400 Hz, Demeaned, LP: 5 Hz |
VAF>90% | 2 | - | - | CP ≠ AD |
| Shuman et al. (2017) | CP: 113 TD: 73 |
CP: I: 10.4 ± 4.8, II: 10.9 ± 5.8, III: 12.2 ± 9.4 TD: 10.3 ± 3.5 |
Spastic (bi) | I, II, III | NG | 5 bi (= 10) |
HP: 40 Hz, LP: 4, 6, 8, 10, 20, 30, 40 Hz |
VAF>90% VAF1 Walk-DMC Different LP cut-offs |
LP 4 Hz
CP: 2.1 ± 0.6 TD: 2.9 ± 0.4 LP 40 Hz CP: 2.9 ± 0.4 TD: 3.4 ± 0.5 |
LP 4 Hz
CP: I = 80†, II = 84†, III = 88† TD: 72 LP 40 Hz CP: I = 75†, II = 78†, III = 82† TD: 62.4 |
LP 4 Hz
CP: I = 82†, II = 75†, III = 64† TD: 100 LP 40 Hz CP: I = 84†, II = 77†, III = 67† TD: 100 |
- |
| Steele et al. (2019) | CP: 20 | 10.4 (6.2–13.6) | Spastic (bi) | I, II, III | >3 | 5 bi (= 10) |
HP: 25 Hz, LP:10 Hz |
VAF>95% VAF1 |
3.1 (range 2–4) | 81.4 ± 5.5 | - | - |
| Shuman et al. (2019) | CP: 147 TD: 31 |
CP: BoNT-A: 6.8 ± 2.9, SDR: 9.3 ± 2.0, SEMLS: 12.1 ± 3.1 TD: 9.3 ± 2.8 |
Spastic (33 uni, 144 bi) |
I, II, III | NG | 8 bi (= 16) |
HP: 20 Hz, LP: 10 Hz |
VAF>90% VAF1 Walk-DMC |
CP: 2.8 ± 0.6 TD: 4.2 ± 0.4 |
CP (pre-treatment): BoNT-A: 79.1 ± 6.2, SDR: 80.1 ± 4.9, SEMLS: 80.2 ± 5.9 TD: 64.4 ± 3.1 |
Improved post-treatment |
Spatial & temporal: Pre-treatment CP ≈ TD |
| Oudenhoven et al. (2019) | CP: 36 | 7.2 (4–13) | Spastic (bi) | I, II, III | 3 | 5 bi (= 10) |
HP: 20 Hz, LP: 2 Hz |
VAF>90% | Higher N = better treatment outcomes | No correlation with treatment outcomes | - | - |
| Kim Y. et al. (2018) | CP: 20 TD: 8 |
CP: 12.5 ± 3.3 TD: 12.0 ± 2.6 |
Spastic (17 uni, 3 bi) |
I, II | 5 (NMF on each stride separately) | 8 bi (= 16) |
HP: 35 Hz, LP: 5 Hz |
VAF>90% VAF1 Walk-DMC |
Mean per stride CP: 3.4 ± 0.3 TD: 3.8 ± 0.2 |
CP: 71 ± 4 TD: 61 ± 3 |
CP: 65 ± 14.2 (40.2–91.3) TD: 100 ± 10 (85.1–113.0) |
Spatial: CP = TD Temporal: CP ≠ TD |
| Steele et al. (2015) | CP: 549 TD: 84 |
CP: 9.8 (7.4–13.3)# TD: 10.3 (7.6–13.0)# |
Spastic (122 uni, 427 bi) |
I, II, III, IV | 1 | 5 bi (= 10) |
BP: 20–400 Hz, LP: 10 Hz |
VAF>90% | CP: >80% = 1 or 2 TD: >60% = 3 |
CP: 84.2 (83.7–84.7) TD: 74.6 (71.3–76.1) |
CP: 86.2 (85.5–86.9) TD: 100 (97.9–102.1) |
Spatial: CP = TD Temporal: CP ≠ TD |
| Shuman et al. (2016) | CP: 5 TD: 6 |
CP: 10.2 (6.0–13.0) TD: 10.3 (6.0–13.0) |
Spastic (2 uni, 3 bi) |
I | CP: 47.5 ± 19.6 (24–81) TD: 44.8 ± 15.9 (25–78) |
8 bi (= 16) |
HP: 40 Hz, LP: 4 Hz |
VAF1 | - | CP: 77.2 ± 4.1 TD: 68.4 ± 2.3 |
- | - |
| Goudriaan et al. (2018) | CP: 15 TD: 15 DMD: 15 |
CP: 8.9 (7.6–9.8)# TD: 8.6 (7.3–10.0)# DMD: 8.7 (6.8–9.9)# |
Spastic (8 uni, 7 bi) |
I, II | 10 | 8 bi (= 16) |
BP: 20–450 Hz, LP: 10 Hz |
VAF1 | - | CP: 74 TD: 65 DMD: 60 |
- | - |
| Schwartz et al. (2016) | CP: 473 | 7.7 ± 3.3 | NG | I, II, III | >4 | 8 bi (= 16) |
NG | Walk-DMC | - | - | Higher walk-DMC pre- treatment = better outcomes | - |
| Shuman et al. (2018) |
Centre 1
CP: 473 TD: 84 Centre 2 CP: 163 TD: 12 |
Centre 1
CP: 7.5 ± 3.4 Centre 2 CP: 9.3 ± 2.7 |
NG | I, II, III | NG |
Centre 1
8 bi (= 16) Centre 2 4 bi (= 8) |
HP: 20 Hz, LP: 10Hz |
Walk-DMC | - | - | CP < TD Higher walk-DMC pre- treatment = better outcomes |
- |
CP, cerebral palsy; TD, typically developing; AD, adults; DMD, duchenne muscular dystrophy; Uni, unilateral; Bi, bilateral; Dysk, dyskinetic; GMFCS, gross motor function classification system; N, number of synergies; HP, high-pass filter, LP, low-pass filter; NMF, non-negative matrix factorization; VAF, variance accounted for; VAF1, variance accounted for by one synergy (%); RMSE, root mean square error; Walk-DMC, dynamic motor control index during walking; SCA, synergy comprehensive assessment; NG, not given; BoNT-A, Botulinum Toxin Type A; SDR, Selective Dorsal Rhizotomy; SEMLS, Single-Event Multilevel Surgery. Age and number of strides are given as mean (± 1 standard deviation or the range when provided by the authors) unless marked by a #, as this signifies the median (25th-75th percentile).
*
Values in figure and text are not in agreement, so these values are extracted from the figure;
†
Signifies that values are extracted from graphical representations and are not precise.