Article Figures & Data
Figures
- Figure 1.
Experimental setup. A, B, Participants grasped the bar around the border between the black and white parts, and rotated it to align the white part with the red LED. On each trial, subjects decided how to grasp the bar: their thumb could be on the white or on the black end of the bar (toward and away grips, respectively). At trial onset, an electronic shutter (B, in white, brown frame) allowed the participant to see the bar and the target LED. A single TMS pulse was delivered either early (100–300 ms) or late (300–500 ms) during the planning phase (A, gray blocks on time line). When the participant moved his right hand from a home key (B, in blue), the electronic shutter became opaque, preventing vision of the hand, of the bar, and of the target LED. C, The participant kept his chin against a chin rest, and his head position relative to the TMS coils was continuously monitored with a video-based frameless stereotaxic system. D, The brain location of the TMS targets (EBA, in blue; IPS, in red) relative to the closest scalp location (in cyan and yellow, respectively). On each trial, TMS was delivered only when the TMS coils were positioned within 5 mm from the desired scalp locations.
- Figure 2.
Psychometric analysis of grip choice. A psychometric procedure quantified the probability of selecting one of two possible grips [p (toward grip)] as a function of the orientation of the bar (in degrees; 0° corresponds to an upward pointing bar/12 o’clock position, with the number of degrees increasing clockwise). A, B, Single-trial choices (dots) were summarized in a grip choice profile with a moving average (A, dashed line) and parameterized with a psychometric function (A, B, green line) given by the sum of two logistic curves, resulting in six parameters (B): an amplitude parameter (∝) capturing to what extent grip choice is influenced by the expected body posture toward the end of the action (posture bias); an offset parameter (β) capturing biases toward either grip type, irrespective of the bar orientation (systematic bias); two slope parameters (γ1, γ2) capturing the range of bar orientations over which participants switch their grip preference from one type to the opposite grip type (switch range); and two phase-offset parameters (δ1, δ2) capturing the orientations at which participants switch their grip preference from one type to the opposite grip type. C, This panel illustrates how trials with a bar orientation evoking equally mixed grip preferences [p (toward grip) = ∼0.5] were classified as MULTIPLE-option (red line). The cutoff value for this category of trials (gray interval) was estimated separately for each participant and included 25% of the trials. The remaining trials evoked consistent grips and were classified as SINGLE option.
- Figure 3.
Effect of TMS on grip choice. Top, Example fits of grip choice of a single representative participant for action-type levels HOLD (left) and ROTATE (right), collapsed over levels of tms-site and tms-time. Dots represent single-trial grip choices (toward, away) after shifting by cosine phase (see Measures of task performance). Single-trial data are limited to 200 randomly selected trials per plot. Bottom, Posture bias (∝) parameters for action-type levels HOLD (left) and ROTATE (right) according to the psychometric function described in Measures of task performance. Error bars indicate standard error of the mean.
- Figure 4.
Effect of TMS on goal-state error. Absolute differences (in degrees) between the desired orientation of the bar (as indicated by the target LED) and the final orientation of the bar for HOLD actions (left column) and ROTATE actions (right column) with SINGLE (top row) and MULTIPLE (bottom row) grip-options. Error bars indicate standard error of the mean.
- Figure 5.
Time-resolved analysis of TMS effects on goal-state error. Top, Temporal dynamics of effects of TMS over EBA (blue) and IPS (orange) on goal-state error, time locked to trial onset (left) and wrist movement onset (right) for MULTIPLE-option ROTATE actions. Bold line sections indicate temporal clusters in which TMS over EBA/IPS had a significantly larger effect than sham stimulation at the same time. Abrupt transitions between datapoints of the time-resolved average are a consequence of the different number of participants contributing to different datapoints. Wrist movement onset refers to the earliest detectable sign of arm motion, namely, changes in wrist position (measured with a motion-tracking system) that occurred systematically earlier than the release of the home button. There were no TMS pulses delivered after participants released the home button. Bottom, Distribution of button press times and wrist movement-onset times (left) or trial-onset times (right) relative to TMS time in the corresponding top panel.
Tables
- Table 1:
Fitted parameters (mean ± SD) and statistical differences (t values, p values) comparing the effects of HOLD and ROTATE conditions on planning time and goal-state error, as well as grip choice (according to the psychometric function described in the Materials and Methods subsection Measures of task performance)
Effects of action type on task performance HOLD trials ROTATE trials HOLD vs ROTATE Planning time (ms) 420 (20) 411 (18) t(21) = 2.471, p = 0.022 Goal-state error (°) 2.96 (1.14) 8.35 (3.26) t(21) = 7.250, p < 0.001 Grip choice Posture bias (∝) 0.967 (0.029) 0.671 (0.244) t(21) = 5.859, p < 0.001 Systematic bias (β) 0.021 (0.022) 0.235 (0.184) t(21) = 5.554, p < 0.001 Switch range (1/γ) 0.092 (0.014) 0.141 (0.076) t(21) = 2.876, p = 0.009 Phase difference (Δφ) 135.4 (71.8) 162.3 (18.5) t(21) = 1.929, p = 0.067
In this issue
