Comparison of EEG-Features and Classification Methods for Motor Imagery in Patients with Disorders of Consciousness

Abstract

Current research aims at identifying voluntary brain activation in patients who are behaviorally diagnosed as being unconscious, but are able to perform commands by modulating their brain activity patterns. This involves machine learning techniques and feature extraction methods such as applied in brain computer interfaces. In this study, we try to answer the question if features/classification methods which show advantages in healthy participants are also accurate when applied to data of patients with disorders of consciousness. A sample of healthy participants (N = 22), patients in a minimally conscious state (MCS; N = 5), and with unresponsive wakefulness syndrome (UWS; N = 9) was examined with a motor imagery task which involved imagery of moving both hands and an instruction to hold both hands firm. We extracted a set of 20 features from the electroencephalogram and used linear discriminant analysis, k-nearest neighbor classification, and support vector machines (SVM) as classification methods. In healthy participants, the best classification accuracies were seen with coherences (mean = .79; range = .53−.94) and power spectra (mean = .69; range = .40−.85). The coherence patterns in healthy participants did not match the expectation of central modulated -rhythm. Instead, coherence involved mainly frontal regions. In healthy participants, the best classification tool was SVM. Five patients had at least one feature-classifier outcome with p0.05 (none of which were coherence or power spectra), though none remained significant after false-discovery rate correction for multiple comparisons. The present work suggests the use of coherences in patients with disorders of consciousness because they show high reliability among healthy subjects and patient groups. However, feature extraction and classification is a challenging task in unresponsive patients because there is no ground truth to validate the results.

Substitute or complement? Defining the relative place of EEG and fMRI in the detection of voluntary brain reactions

“…For example, in the sample of 14 patients recruited by Gibson et al (2014), some of them were ineligible for the fMRI and/or the EEG evaluation and only six could complete both procedures. Promising results have recently been found in EEG as well as in fMRI (Schnakers et al, 2009;Goldfine et al, 2011;Cruse et al, 2012b;Gibson et al, 2013;Holler et al, 2013;Naci and Owen, 2013). Better sensitivity may be observed with these protocols than with the one used in the present study, but since these protocols were performed by different teams in different groups of subjects it is difficult to determine precisely.…”

“…Indeed, in this protocol, the cross-validation design is a particularly delicate issue that has already been pointed out (Goldfine et al, 2013;Noirhomme et al, 2014) and that shows the strong influence of the temporal dependence between the test-set blocks on the classification accuracy. In recent studies avoiding this limitation, promising results have been found in healthy volunteers but also in some patients with DOCs (Gibson et al, 2013;Holler et al, 2013) but their replication on a larger scale is needed before clearly establishing whether they are sensitive enough to reliably capture awareness. By contrast, the conservatism of FDR correction is more a matter of debate.…”

“…As a consequence, several research groups have begun to focus more on electroencephalography (EEG), which is less expensive and transportable to the patient bedside. With protocols similar to fMRI, some EEG studies showed that patients with DOCs were able to willfully direct their attention toward specific stimuli (Schnakers et al, 2008;Bekinschtein et al, 2009;Chennu et al, 2013) or to perform mental imagery tasks (Cruse et al, 2011(Cruse et al, , 2012aGoldfine et al, 2011;Holler et al, 2013). For example, in the experiment from (Cruse et al, 2011(Cruse et al, , 2012a, patients with DOCs were asked to imagine either squeezing their right hand or wiggling their toes.…”

Correlation of EEG spectra, connectivity, and information theoretical biomarkers with psychological states in the epilepsy monitoring unit — A pilot study

“…The presented measures were shown in the past to be potential biomarkers for the classification of disorders of consciousness [43]. Brainrate was intended to serve as a standard indicator of activation and mental arousal [19], and was found to be indicative for attention deficit hyperactivity disorder [44].…”

Review of Brain Computer Interface Components and Corresponding Research Topics

“…Several papers compare the performance of classifiers to provide a guide for building a BCI system [134][135][136][137][138].…”