Using Transcranial Magnetic Stimulation to Test a Network Model of Perceptual Decision Making in the Human Brain

Previous research has suggested that the lateral occipital cortex (LOC) is involved with visual decision making, and specifically with the accumulation of information leading to a decision. In humans, this research has been primarily based on imaging and electroencephalography (EEG), and as such only correlational. One line of such research has led to a model of three spatially distributed brain networks that activate in temporal sequence to enable visual decision-making. The model predicted that disturbing neural processing in the LOC at a specific latency would slow object decision-making, increasing reaction time (RT) in a difficult discrimination task. We utilized transcranial magnetic stimulation (TMS) to test this prediction, perturbing LOC beginning at 400 ms post-stimulus onset, a time in the model corresponding to LOC activation at a particular difficulty level, with the expectation of increased RT. Thirteen healthy adults participated in two TMS sessions in which left and right LOC were stimulated separately utilizing neuronavigation and robotic coil guidance. Participants performed a two-alternative forced-choice task selecting whether a car or face was present on each trial amidst visual noise pre-tested to approximate a 75% accuracy level. In an effort to disrupt processing, pairs of TMS pulses separated by 50 ms were presented at one of five stimulus onset asynchronies (SOAs): -200, 200, 400, 450, or 500 ms. Behavioral performance differed systematically across SOAs for RT and accuracy measures. As predicted, TMS at 400 ms resulted in a significant slowing of RT. TMS delivered at -200 ms resulted in faster RT, indicating early stimulation may result in priming and performance enhancement. Use of TMS thus causally demonstrated the involvement of LOC in this task, and more broadly with perceptual decision-making; additionally, it demonstrated the role of TMS in testing well-developed neural models of perceptual processing.