摘要Objective Behavioral studies have suggested a low-frequency (0.05 Hz) fluctuation of sustained attention on the basis of the intra-individual variability of reaction-time.Conventional task designs for functional magnetic resonance imaging (fMRI) studies are not appropriate for frequency analysis.The present study aimed to propose a new paradigm,real-time finger force feedback (RT-FFF),to study the brain mechanisms of sustained attention and neurofeedback.Methods We compared the low-frequency fluctuations in both behavioral and fMRI data from 38 healthy adults (19 males; mean age,22.3 years).Two fMRI sessions,in RT-FFF and sham finger force feedback (S-FFF) states,were acquired (TR 2 s,Siemens Trio 3-Tesla scanner,8 min each,counter-balanced).Behavioral data of finger force were obtained simultaneously at a sampling rate of 250 Hz.Results Frequency analysis of the behavioral data showed lower amplitude in the lowfrequency band (0.004-0.104 Hz) but higher amplitude in the high-frequency band (27.02-125 Hz) in the RT-FFF than the S-FFF states.The mean finger force was not significantly different between the two states.fMRI data analysis showed higher fractional amplitude of low-frequency fluctuation (fALFF) in the S-FFF than in the RT-FFF state in the visual cortex,but higher fALFF in RT-FFF than S-FFF in the middle frontal gyrus,the superior frontal gyrus,and the default mode network.Conclusion The behavioral results suggest that the proposed paradigm may provide a new approach to studies of sustained attention.The fMRI results suggest that a distributed network including visual,motor,attentional,and default mode networks may be involved in sustained attention and/or real-time feedback.This paradigm may be helpful for future studies on deficits of attention,such as attention deficit hyperactivity disorder and mild traumatic brain injury.