Single-session communication with a locked-in patient by functional near-infrared spectroscopy

Abstract

There is a growing interest in the possibility of using functional neuroimaging techniques to aid in detecting covert awareness in patients who are thought to be suffering from a disorder of consciousness. Immerging optical techniques such as time-resolved functional near-infrared spectroscopy (TR-fNIRS) are ideal for such applications due to their low-cost, portability, and enhanced sensitivity to brain activity. The aim of this case study was to investigate for the first time the ability of TR-fNIRS to detect command driven motor imagery (MI) activity in a functionally locked-in patient suffering from Guillain-Barré syndrome. In addition, the utility of using TR-fNIRS as a brain-computer interface (BCI) was also assessed by instructing the patient to perform an MI task as affirmation to three questions: (1) confirming his last name, (2) if he was in pain, and (3) if he felt safe. At the time of this study, the patient had regained limited eye movement, which provided an opportunity to accurately validate a BCI after the fNIRS study was completed. Comparing the two sets of responses showed that fNIRS provided the correct answers to all of the questions. These promising results demonstrate for the first time the potential of using an MI paradigm in combination with fNIRS to communicate with functionally locked-in patients without the need for prior training.

Keywords: Guillain–Barré syndrome; brain–computer interface; functional near-infrared spectroscopy; motor imagery; time-resolved measurements.

Fig. 1

Schematic of the TR-fNIRS probes on the head. The red circle illustrates the location of the emission fiber (FCz), whereas the blue circles represent the detection fiber positions with a source–detector distance of 3 cm.

Fig. 2

Schematic of the MI paradigm used to communicate with the patient. Following 30 s of rest period, each question was repeated five times in a alternating block design of task and rest period for a total experiment time of 5:30 minutes per question. During the task period, the patient was instructed to imagine playing tennis if he wanted to answer “yes” or to stay relaxed if he wanted to answer “no.”

Fig. 3

Changes in the concentration of oxyhemoglobin (red) and deoxyhemoglobin (blue) averaged across all five cycles for each of the three questions. For the responses classified as “yes” (i.e., correct last name and do you feel safe), the signals were averaged across all activated channels, whereas for the response classified as “no” (i.e., are you in pain), the signals were averaged across all four channels. The baseline time course labeled “rest” refers to data acquired without MI activation and is presented as a reference for the contrast observed during the question periods. The error bars represent the standard error of mean across the specific activated/inactivated channels. The gray boxes indicate the response period.