Table 2.
Nonlinear EEG measurements in disorders of consciousness patients
| Clinical application | Literature | Non-linear measurements | Subjects (numbers) | Main findings |
|---|---|---|---|---|
| Diagnosis | Gosseries et al. [32] | State entropy and response entropy |
HC 16 MCS 26 VS/UWS 24 Coma 6 |
EEG entropy correlated with CRS-R total scores Mean EEG entropy values were higher in MCS than in VS/UWS |
| Sara et al. [29] | ApEn |
HC 40 VS/UWS 38 |
Mean ApEn values were lower in patients than in HC | |
| Wu et al. [28] | LZC, ApEn |
HC 30 MCS 16 VS/UWS 21 |
VS/UWS had the lowest LZC and ApEn, followed by MCS, and HC had the highest | |
| Sitt et al. [13] | PE, KCC |
HC 14 CS 24 MCS 68 VS/UWS 75 |
PE-based measures were particularly efficient in the theta frequency range, discriminating VS/UWS from other groups KCC significantly discriminated VS/UWS from MCS, particularly for electrodes over parietal region |
|
| Piarulli et al. [14] | Spectral entropy |
MCS 6 VS/UWS 6 |
MCS showed higher spectral entropy mean value and higher time variability MCS were characterized by spectral entropy fluctuations with periodicities at 70 min, the periodicities closely resemble those described in awake HC |
|
| Thul et al. [40] | PE |
HC 24 MCS 7 VS/UWS 8 |
VS/UWS showed considerably reduced PE compared to MCS Overall differences of PE were significant between HC and MCS, and between HC and VS/UWS |
|
| Stefan et al. [19] | ApEn and PE |
MCS 11 VS/UWS 51 |
ApEn in all frequency ranges was higher in MCS than VS/UWS PE in the alpha range was significantly higher in MCS than VS/UWS |
|
| Engemann et al. [30] | PE, KCC and Spectral entropy |
MCS 179 VS/UWS 148 |
PE was one of the top ranked among > 20 potential biomarkers for classification of of DOC patients |
| Clinical application | Literature | Non-linear measurements | Subjects (numbers) | Follow-up times | Main findings |
|---|---|---|---|---|---|
| Prognosis | Sara et al. [29] | ApEn |
HC 40 VS/UWS 38 |
6 months | Patients with the lowest ApEn values either died (n = 14) or remained in a VS/UWS (n = 12), whereas patients with the highest ApEn values became MCS (n = 5) or showed partial (n = 4) or full recovery (n = 3) |
| Wu et al. [28] | LZC, ApEn |
CS 30 MCS 16 VS/UWS 21 |
6 months | LZC and ApEn under painful stimulation increased more significantly in patients who recovered than in patients without recovery | |
| Stefan et al. [19] | ApEn, PE |
MCS 11 VS/UWS 51 |
589 ± 1125 days | PE in the delta, theta bands predicted outcome better than ApEn |
| Clinical application | Literature | Non-linear measurements | Subjects (numbers) | Treatments | Main findings |
|---|---|---|---|---|---|
| Treatment evaluation | Wang et al. [31] | PE |
MCS 7 VS/UWS 7 |
SCS | PE increased as compared to the baseline during SCS turning on |
| Hermann et al. [23] |
PE, KCC, Spectral entropy |
EMCS 4 MCS 32 VS/UWS 24 |
tDCS | PE in the theta-alpha band showed a trend towards an increase in responders as compared to non-responders in the same parietal region |
HC healthy control, DOC disorders of consciousness, MCS minimally conscious state, VS/UWS vegetative state/unresponsive wakefulness state, CS conscious patients, BIS bispectral index, ApEn approximate entropy, LZC Lempel–Ziv Complexity, PE permutation entropy, KCC Kolmogorov-Chaitin complexity, SCS spinal cord stimulation, tDCS transcranial direct current stimulation