Abstract
BACKGROUND: Behavioral assessment based on external manifestations of consciousness fails for patients with cognitive motor dissociation (CMD). Functional near-infrared spectroscopy (fNIRS) is an emerging neuroimaging technique that can detect internal brain functional activities. However, the extent to which fNIRS can help identify CMD patients among those with disorders of consciousness (DOC) remains unclear. OBJECTIVE: To identify CMD patients among DOC patients using fNIRS with a command-driven hand-open-close motor imagery task. METHODS: fNIRS was used to measure the hemodynamic responses of 70 prolonged DOC patients, including 30 with vegetative state/unresponsive wakefulness syndrome (VS/UWS), 20 with minimally conscious state minus (MCS-), and 20 with minimally conscious state plus (MCS+), during a command-driven hand-open-close motor imagery task. Seven features of hemodynamic responses were extracted during the task and the rest conditions. The support vector machine combined with genetic algorithm was employed to classify and predict the brain's response to spoken commands and to identify CMD patients among prolonged DOC individuals. RESULTS: We identified seven CMD patients using fNIRS, of whom four were in VS/UWS and three were in MCS-. Six months after fNIRS examination, the seven identified CMD patients were more likely to have a favorable outcome (3/4 vs. 1/31, P = 0.014, Fisher's exact test) compared with non-CMD patients. CONCLUSIONS: CMD patients can be identified through fNIRS combined with a command-driven motor imagery task, which will aid in the accurate diagnosis of DOC patients.