Complete Locked-in and Locked-in Patients: Command Following Assessment and Communication with Vibro-Tactile P300 and Motor Imagery Brain-Computer Interface Tools.

Vincenzo La Bella,Christoph Guger,Rupert Ortner,Brendan Z Allison,Rossella Spataro,Woosang Cho,Alexander Heilinger

doi:10.3389/fnins.2017.00251

Vincenzo La Bella, Christoph Guger + Show 5 more

Open Access

https://doi.org/10.3389/fnins.2017.00251

Copy DOI

Abstract

Many patients with locked-in syndrome (LIS) or complete locked-in syndrome (CLIS) also need brain-computer interface (BCI) platforms that do not rely on visual stimuli and are easy to use. We investigate command following and communication functions of mindBEAGLE with 9 LIS, 3 CLIS patients and three healthy controls. This tests were done with vibro-tactile stimulation with 2 or 3 stimulators (VT2 and VT3 mode) and with motor imagery (MI) paradigms. In VT2 the stimulators are fixed on the left and right wrist and the participant has the task to count the stimuli on the target hand in order to elicit a P300 response. In VT3 mode an additional stimulator is placed as a distractor on the shoulder and the participant is counting stimuli either on the right or left hand. In motor imagery mode the participant is instructed to imagine left or right hand movement. VT3 and MI also allow the participant to answer yes and no questions. Healthy controls achieved a mean assessment accuracy of 100% in VT2, 93% in VT3, and 73% in MI modes. They were able to communicate with VT3 (86.7%) and MI (83.3%) after 2 training runs. The patients achieved a mean accuracy of 76.6% in VT2, 63.1% in VT3, and 58.2% in MI modes after 1–2 training runs. 9 out of 12 LIS patients could communicate by using the vibro-tactile P300 paradigms (answered on average 8 out of 10 questions correctly) and 3 out of 12 could communicate with the motor imagery paradigm (answered correctly 4,7 out of 5 questions). 2 out of the 3 CLIS patients could use the system to communicate with VT3 (90 and 70% accuracy). The results show that paradigms based on non-visual evoked potentials and motor imagery can be effective for these users. It is also the first study that showed EEG-based BCI communication with CLIS patients and was able to bring 9 out of 12 patients to communicate with higher accuracies than reported before. More importantly this was achieved within less than 15–20 min.

Highlights

While earlier results have shown that persons with LIS resulting from ALS can communicate with BCIs that do not rely on visual stimuli, the present study extends these results in three ways. This is the first study that shows that an EEG-based BCI system can establish communication with CLIS
It shows that non-visual BCI technology can be used to assess command following in these patients. This is the first study to show that the mindBEAGLE system, which is designed to be easy to use without technical expertise, can provide a practical and usable platform for both assessment and communication in these patients
Like most BCIs, the trade-off between speed and accuracy is largely a function of each user’s abilities and preferences; the MI and VT3 paradigms could be faster at the expense of reduced accuracy, and vice versa

Summary

Introduction

Minimal consciousness (MCS) patients do not have reliable voluntary motor control, and seem to have substantial cognitive impairment, their cognitive function may fluctuate. These three types of patients are typically categorized as having a disorder of consciousness (DOC). Locked-in (LIS) and completely locked-in (CLIS) patients show limited or no motor response, and assessing cognitive function can be difficult. LIS and CLIS patients may have healthy cognitive function, or may exhibit substantial impairment (Kübler et al, 2001)

Methods

Results

Discussion

Conclusion