Additional file 1.

Comparison of individual activation maps obtained from our 5 minutes protocol with maps computed from fMRI data acquired during the same session with another longer bloc design. As explained in the manuscript, our 5 minute protocol was added to each fMRI session performed in our laboratory. To estimate efficiency of our fast functional mapping and see if its results should be considered as free from its particular design, we compared some of our statistical maps with those corresponding to similar conditions performed in the associated main protocol (see details of each protocol below). We display here two examples of individual contrasts for three different fMRI protocols: one protocol including visual sentence reading (p = 0.001 corrected at the voxel level), and two other protocols including each a calculation task and a verbal control task (p = 0.001 uncorrected at the voxel level). Surrounded by a blue frames are reported a statistical map form the 5 min protocole for the corresponding subject at a similar anatomical location (p = 0.01 uncorrected at the voxel level, 30 voxels for cluster extent). Note that subject 2 performed twice our 5 min protocol with an interval of nine weeks, illustrating reliability of the activation topography. These maps suggest that even with quite different experimental conditions (block design versus fast event-related design, different control tasks, different rate of presentation, sometime different notation, different resolutions, different number of trials...) our 5 min design was able to capture most of the individual cerebral sites that characterize subject's functional activation during a task (here reading or calculating). In conclusion, maxima reported for each subject could be reasonably associated to areas that are crucial to perform a specified cognitive task, independently of the experimental conditions of stimulation and acquisition. Statistical thresholds were adapted for the obvious reason that the statistical power of a protocol depends on the number of trials performed. Experiment 1. Three sessions comprising each 20 miniblocks of 4 visual sentences (total duration = 25 mn/240 sentences). The sentences varied in length from 5 to 17 words, flashed successively at a rate of one word every 270 msec. Subjects had just to read passively sentences and press a button when a probe sentence was displayed (6% of trials). Data acquisition was similar to our protocol (TR = 2.4 s, 24 slices, voxel size = 3 × 3 × 5 mm). Experiment 2. Two sessions comprising each 2 miniblocks of mental calculation and 2 miniblocks of a numerical control task. Each miniblock comprised 9 mental subtractions (total duration = 9 mn/36 trials). Subject had to subtract a visually displayed one Arabic digit number from a memorized reference ('12' for instance) and silently pronounced the result. Numbers was flashed for 200 ms and successively at a rate of one digit every 2200 msec. During the control task, the subject had to pronounce silently the numerical successor of the stimuli ('4' for the stimuli '3'). Data acquisition was close to those of our protocol (TR = 2.4 s, 24 slices, voxel size = 3 × 3 × 5 mm). Experiment 3. One session comprising each 5 miniblocks of mental calculation and 5 miniblocks of a non-numerical control task. Each miniblock comprised 10 mental calculations (total duration = 10 mn/50 mental numerical trials). During mental calculation, subject had to make a serie of subtraction and addition: each visually proposed arithmetical operation (like '+3' or '-5') was applied to the result of the previous calculation. After a serie of five operations, the subject indicated by pressing a button of the final memorized result was similar to the result proposed on the video screen. During the control task, subjects saw a flow of letters. After a serie of five letters, he indicated in a similar way if a visually displayed target letter was present in the previous stimuli. Because a high resolution acquisition was used in this paradigm, parameters of acquisition differed much from our protocol: acquisition was limited to a volume placed on superior frontal and parietal lobes (TR = 3 s, 32 slices, voxel size = 1.5 × 1.5 × 2 mm).

Format: TIFF Size: 1.7MB Download file

Pinel et al. BMC Neuroscience 2007 8:91   doi:10.1186/1471-2202-8-91