Multi-variate pattern analysis (MVPA) applied to BOLD-fMRI has verified successful at decoding complicated fMRI signal patterns associated with a variety of cognitive processes. to identify regions that carry sufficient information to permit significantly above-chance prediction of “Yes/No” thoughts assumptions about response shape inherent to most fitting techniques. As previously mentioned trial averaging was used to optimize the signal-to-noise percentage of spatial patterns input to the classifier. If spatial patterns from a given region genuinely reflect reactions evoked from the “Yes/No” thoughts then decoding accuracy should increase towards desirable goal of 100% as signal-to-noise ratio increases. If not decoding accuracy should not change with the number of averaging. We tested this working hypothesis in all regions marked as useful in Experiment 1 as well as in a SKLB1002 control region not expected to carry robust predictive information (see below). We also performed a permutation test in which we randomized the “Yes/No” labels of the trials being averaged. This provided a second control condition to compare against. Participants Seven healthy volunteers (2 females age range 22-43 years mean age = 29) were recruited at National Institute of Mental Health Bethesda MD USA. All participants had normal or corrected-to-normal vision and had no history of neurological MAP2K2 or psychiatric disorder. After a full description of the study written informed consent was obtained from each participant following a protocol approved by the Ethical Committee at National Institute of Mental Health USA. SKLB1002 Experimental SKLB1002 Materials The visual stimuli and pre-scan questionnaire for this experiment were presented in English. One hundred and twenty-four “Yes/No” questions about simple facts were used. Most of these questions correspond to English translations of the questions used in Experiment 1. A small set of questions was substituted by completely new questions to fit cultural differences. For half of the questions (62 questions) the expected answer was “Yes”. Experimental Paradigm In this second experiment the paradigm from Experiment 1 was altered to become a “slow” event-related design with constant long delay periods (12s). Thus the regression step is no longer needed and hemodynamic responses for each trial type can be obtained by simple trial averaging (point-to-point averaging of the signal after the question onset). This paradigm is usually shown in Fig. 2. Each trial had the following structure: intentional cue (2s); simple-fact question (4s); delay period (8s); response prompt (2.3s); fixation period (11.7s). The intentional cue consisted of the word “Honest” written in a green font to instruct subjects to respond honestly or the word “Dishonest” written in a red font to instruct subjects to respond dishonestly. During the question period the question appeared in the center of the screen in white font. During the delay period a fixation cross appeared at the center of the screen and subjects were instructed to keep their answer in mind. For the prompt to respond one of two possible texts appeared in the center of the screen: “Y N” which instructed participants to use the button box in their left hands to answer “Yes” and the one in their right hands to answer “No”; or “N Y” which corresponded to the opposite assignment of “Yes/No” to the right/left hands. Which prompt screen appeared at the end of each trial was randomized. This randomization precluded subjects from keeping SKLB1002 their responses in mind during the delay period in terms of motor actions as they don’t know which button signals “Yes” and which button signals “No” until the end of the delay period. The 124 questions (62 expecting a “Yes” and 62 expecting a “No”) were randomly assigned into 6 runs. The intentions (“Honest/Dishonest”) and the two types of response prompts (“Y N” and “N Y”) were assigned to the questions in a counter-balance manner. Physique 2 One trial of experimental paradigm for experiment 2. At the beginning of each trial an training cue is displayed for 2s asking subjects to be either honest or dishonest when answering the following question. Then a question on simple facts is usually presented … Data acquisition A Siemens 7T scanner equipped with a 32-channel head coil was used in Experiment 2. Functional scans SKLB1002 were obtained using a gradient-echo EPI sequence (54 slices TR/TE = 2000/25ms slice thickness = 2mm FOV = 192 mm flip angle = 50° image matrix: 96×96 343 volumes) covering the whole brain. An additional high-resolution anatomical scan was acquired in.