BCIwiki (bciwiki.com) — Researchers at the Fourth Military Medical University published an fNIRS dataset acquired during a multi-day Stroop task, comprising frontal hemoglobin responses from 55 young adults, providing a new optical neuroimaging resource for cognitive state decoding in brain-computer interfaces. The study was published online on August 6, 2026, in Scientific Data, by authors Zeng L, Sun K, Yuan Y, Gao Y, Wang X, and Wen Z, affiliated with the Department of Medical Psychology and the Department of Aerospace Medicine at the Fourth Military Medical University.
The dataset was collected from 55 young adults, each completing three sessions of a color-word Stroop task within approximately 2 weeks, with more than 30 trials per condition to avoid mental fatigue caused by a single session. The team noted that datasets for decoding cognitive states using optical neuroimaging modalities remain scarce. The dataset supports applications ranging from studying conflict inhibition to building decoders for neurofeedback training and large-scale cross-subject fNIRS models, as well as facilitating the development of signal processing algorithms for hemodynamic signals.