Methods: roi analyses



Download 17.51 Kb.
Date conversion28.12.2016
Size17.51 Kb.
Supplementary Materials
Methods:

ROI analyses:

As we have a sample of n = 1 for DD, the validity of a group X condition ANOVA may be questionable. Nevertheless, as an exploratory analysis, we investigated whether the MNS and Mentalizing ROIs were modulated by group by performing a 2 (Group: DD/TD) x 2 (Condition: Possible/Impossible) ANOVAs for each ROI (percent BOLD signal change as dependent variable).

We furthermore conducted a post-hoc analysis to visualize where D.D. ROI responses fell in relation to the estimated sample distribution for TD subjects. We estimated the 95% CI of the sample distribution as mean +/- 1.65*sample standard deviation. We quantified these observations by calculating the percentile score of DD ROI activity in relation to the TD sample distribution. We note that comparing a second level analysis in TD against a first level analysis in DD is statistically challenging, so our comparisons are for descriptive purposes only.



Whole brain analyses:

Action Observation: Given D.D.’s atypical physical body, it is probable that her motor and sensory cortices are developed differently from TD individuals. For this reason, as well as to observe D.D.'s brain activity beyond the a priori ROIs, we also conducted whole-brain analyses when D.D. observed possible and impossible actions (“all possible actions – controls,” “all impossible actions – controls”). We also assessed the common representations of possible and impossible actions by performing a conjunction analysis ("all possible actions - controls", "all impossible actions - controls", each thresholded at p<0.05, FDR, k>5). Finally, we directly compared observation of possible actions against observation of impossible actions (“all impossible actions – all possible actions”, "all possible actions - all impossible actions").

Pain Observation: To observe D.D.'s brain activity beyond a priori ROIs, we also looked at the whole-brain activation patterns when D.D. observed pain versus no pain video clips, both when viewing pain in a body part that she has (“mouth pain – mouth no pain”) as well as when viewing a body part that she does not have (“hand pain – hand no pain”). To investigate differences in how D.D. processes pain to a body part she lacks versus pain to a body part she has, we also directly compared D.D.'s brain activity when she observed pain to the hand versus pain to the mouth (“hand pain – mouth pain” and “mouth pain – hand pain”).
Results:

As DD is a single subject compared to a group, we aimed to maximize the amount of data we were able to obtain from her. Thus DD completed two sessions of the study with slightly longer run times. The control subjects instead completed one session of the study. Post hoc analyses indicate that including only half the number of trials for DD does not significantly change the results. That is, even when analyzing only four runs from DD’s data, the percent BOLD signal change is significantly higher than the signal change in TD. Furthermore, we found no significant difference in the overall effect sizes in the ROIs between DD and TD (p=.59).



ROI ANOVA analyses: In the MNS ROIs, we find a main effect of group, with DD showing greater percent BOLD signal change than TD participants in the following ROIs: L IPL, R IPL, L pMC, R pMC, and precuneus (p<0.05 for each). A marginal main effect for group was found in L TPJ (p=0.06). No main effect for condition, or group by condition interactions were found in any ROIs.

Comparison to the normal sample distribution: When visualizing DD ROI responses in relation to the estimated sample distribution for TD subjects, we find that although DD ROI activations were within the 95% CI of the sample distribution, DD seems to recruit certain ROIs to a greater degree than TD subjects, in particular the mPFC when observing actions impossible for herself and S2 when observing another person experience pain in a body part she does not have. We quantified these observations by calculating the percentile score of DD ROI activity in relation to the TD sample distribution (Figure S5).
Whole brain analyses:

Action Observation:

Observation of all possible actions – controls: In D.D., the contrast “all possible actions – controls” revealed activations in regions associated with the putatitive mirror system (bilateral inferior precentral gyri/posterior pars opercularis, bilateral parietal cortices (from superior postcentral gyrus extending ventrally into inferior parietal lobule and parietal operculum), bilateral posterior middle temporal gyri (extending posteriorly into temporal-occipital junction and lateral occipital gyri). These findings are consistent with results from the ROI analyses, and are shown in Figure S2. The left insula was also significantly active (p<0.05, FDR, k>5). A comparison between D.D.’s activation patterns and TD’s activation patterns is shown in Figure S3.

Observation of all impossible actions - controls: The contrast “all impossible actions – controls” revealed activation patterns similar to the “all possible actions – controls” contrast. Regions associated with the mirror system were active including the bilateral precentral/posterior inferior frontal gyri, bilateral superior and inferior parietal cortices (including parietal operculum), and bilateral posterior middle temporal gyri (extending posteriorly into temporal-occipital junction and lateral occipital gyri) (p<0.05, FDR, k>5). Additional regions in left middle frontal gyrus and left medial frontal gyrus (including portions of the supplemental motor area) were also active. These results are shown in Figure S2. A comparison between D.D.’s activation patterns and TD’s activation patterns is shown in Figure S3.

Observation of possible and impossible actions: In the conjunction analysis (regions involved in observation of possible and impossible actions), we found activations in medial superior frontal gyri (supplementary motor area), bilateral superior and middle frontal gyri, inferior frontal gyri (opercularis, orbitalis), bilateral precentral gyri, bilateral superior parietal cortices (including parts of superior postcentral gyrus), inferior parietal cortices (angular and supramarginal gyri), bilateral posterior superior temporal gyri (extending inferior and posterior into middle temporal gyri and middle and inferior occipital cortices), middle cingulate gyri (clusters found in anterior and posterior aspects of middle cingulate gyri), bilateral insula, bilateral thalamus, bilateral putamen, and superior/anterior and inferior/posterior portions of the cerebellum.
Observation of possible actions vs impossible actions: In the direct comparison "all possible actions - all impossible actions", no regions were active even at a lenient threshold of p<0.005, uncorrected. In the opposite comparison "all impossible actions - all possible actions", activations were found in left superior frontal gyrus (medial and superior aspects, including mPFC), right middle frontal gyrus (including DLPFC), left inferior frontal gyrus (opercularis and triangularis), right posterior superior temporal gyrus (including TPJ), and right globus pallidus (p<0.005, uncorrected).

Pain Observation



Whole-brain analyses: In D.D., the contrast “mouth pain – mouth no pain” revealed activations in components of the “pain-matrix” including: the right anterior cingulate gyrus, right superior postcentral gyrus (SI), bilateral inferior parietal operculums (SII), and bilateral insula. In addition the left superior and medial frontal gyri, bilateral inferior frontal gyrus, the left precentral gyrus, bilateral posterior medial cortices, left supramarginal gyrus, bilateral anterior superior temporal gyrus, and right posterior middle temporal gyri were found to be active (p<0.05, FDR, k>5). The contrast "hand pain - hand no pain" revealed activations in the left inferior frontal gyrus, left inferior parietal lobule, and right anterior superior temporal gyrus (cluster extends into the right anterior insula) (p<0.05, FDR, k>5; Figure S4).
We also looked at direct comparisons between observation of pain to a body region D.D. has (mouth) as compared to observation of pain to a body region D.D. lacks (hand). The contrast “mouth pain – hand pain” revealed activations in bilateral superior parietal cortices (extending into the postcentral sulcus), consistent with the results from the ROI analyses, as well as regions in the posteromedial cortices, bilateral posterior middle temporal/middle occipital gyri, bilateral cerebellum, right middle frontal gyrus, and right anterior superior frontal gyrus (p<0.05, FDR, k>5). The opposite contrast, “hand pain – mouth pain”, revealed no significant activations at FDR-corrected p<0.05.


Figure S1. (A) MNS, (B) Mentalizing, and (C) Pain Observation ROIs overlaid on an anatomical image of D.D.’s brain (normalized to MNI space).
Figure S2. D.D. Action Observation Whole-brain Contrast Map. Red: Possible Actions - Control Green: Impossible Actions – Control. Yellow indicates regions of overlap. Both contrasts shown at p<0.05, FDR. Active regions include (1) (anterior to posterior) L Precentral Gyrus, L Inferior Parietal Lobule, L Superior Temporal Gyrus, L Posterior Middle Temporal Gyrus (2) (superior to inferior) L Medial Frontal Gyrus, Bilateral Middle Frontal Gyri, Bilateral Precentral Gyri, L Insula (3) (anterior to posterior) R Inferior Frontal Gyrus, R Inferior Parietal Lobule, L Superior Temporal Gyrus, R Posterior Middle Temporal Gyrus (4) (anterior to posterior) Medial Prefrontal Cortex, L Medial Frontal Gyrus/L ACC, L Posteromedial Cortices (5) (anterior to posterior) L Medial Frontal Gyrus/L ACC, L Superior Frontal Gyrus, Bilateral Precentral Gyri, Bilateral Inferior Parietal Lobules (6) (anterior to posterior) R Middle Frontal Gyrus, R Postcentral Gyrus (extending into inferior parietal), R Posterior Middle Temporal Gyrus (extending into temporal-occipital junction and lateral occipital gyrus).
Figure S3. D.D. and TD Action Observation Whole-brain Contrast Maps. (A) Red: Brain regions active when D.D. views Possible Actions - Control (p<0.05, FDR; T > 3 for display purposes). Green: Brain regions active when TD participants view the same subset of actions - Control (T > 3 for display purposes). Yellow shows regions of overlap. (B) Red: Brain regions active when D.D. views Impossible Actions - Control (p<0.05, FDR; T > 3). Green: Brain regions active when TD participants view the same subset of actions - Control (T > 3 for display purposes). Yellow shows regions of overlap.
Figure S4. D.D. and TD Pain Observation Whole-brain Contrast Maps. Red: Brain regions active when D.D. views All Pain - All No Pain stimuli (p<0.05, FDR; T > 3). Green: Brain regions active when TD participants view All Pain - All No Pain stimuli (T > 3 for display purposes). Yellow shows regions of overlap in left secondary sensory cortices, right post-central gyrus and right anterior insular cortices.
Figure S5. (TOP) DD mentalizing ROI activations (impossible - possible actions) are plotted next to the 95% CI of the TD sample distribution (estimated as sample mean +/- 1.65*sample standard deviation). DD mentalizing ROI percentile scores are: 76 (L TPJ), 69 (R TPJ), 92 (mPFC), 76 (precuneus). Although mentalizing ROI activations in DD fall below the 95 %tile of the TD sample distribution, DD activations are generally greater than TD individuals. The joint probability of DD showing greater activation than the median of the TD sample across all mentalizing regions is 0.06. (BOTTOM) DD pain ROI activations (mouth pain - hand pain) are plotted next to the 95% CI of the TD sample distribution (estimated as sample mean +/- 1.65*sample standard deviation). The percentile scores of DD pain ROI activations are: 69 (L S1), 92 (L S2), 46 (L mid insula), 69 (R S1), 53 (R S2), 30 (R anterior insula), 84 (R mid insula), 38 (anterior midbrain), 69 (mid cingulate), 46 (posterior midbrain).


The database is protected by copyright ©dentisty.org 2016
send message

    Main page