Common and distinct neural correlates of social interaction processing and theory of mind in narratives.

1. function [mvmt_matrix mvmt_regs_24, mvmt_table] = movement_regressors(mvmt_file)
2. % Read a realignment parameter file and create a set of movement regressors with quadratic/derivative transformations
3. %
4. % :Usage:
5. % ::
6. % [mvmt_matrix, mvmt_regs_24, mvmt_table] = movement_regressors(mvmt_file)
7. %
8. % :Inputs:
9. %
10. % **mvmt_file:**
11. % String. Full path to the movement parameter file (e.g., rp_sub-sid001567_task-pinel_acq-s1p2_run-03_bold.txt)
12. % generated after realignment (motion correction). This file contains the head motion
13. % estimates (translations and rotations) required to align each image to a reference image.
14. %
15. % :Outputs:
16. %
17. % **mvmt_matrix:**
18. % Numeric matrix. The raw movement parameters imported from mvmt_file.
19. %
20. % **mvmt_regs_24:**
21. % Numeric matrix. A matrix containing 24 movement-related regressors for each image.
22. % These regressors include the standardized movement parameters, their squared values,
23. % first differences (derivatives), and squared first differences.
24. %
25. % **mvmt_table:**
26. % Table object. A table version of mvmt_regs_24 with one variable per regressor.
27. % The variable names are as follows:
28. % - 'mvmt_x', 'mvmt_y', 'mvmt_z', 'mvmt_r', 'mvmt_p', 'mvmt_y'
29. % - 'mvmt_x^2', 'mvmt_y^2', 'mvmt_z^2', 'mvmt_r^2', 'mvmt_p^2', 'mvmt_y^2'
30. % - 'mvmt_xdiff', 'mvmt_ydiff', 'mvmt_zdiff', 'mvmt_rdiff', 'mvmt_pdiff', 'mvmt_ydiff'
31. % - 'mvmt_xdiff^2', 'mvmt_ydiff^2', 'mvmt_zdiff^2', 'mvmt_rdiff^2', 'mvmt_pdiff^2', 'mvmt_ydiff^2'
32. %
33. % :Examples:
34. % ::
35. % % Example: Compute movement regressors from a realignment parameters file.
36. % mvmt_file = fullfile(pwd, 'rp_sub-sid001567_task-pinel_acq-s1p2_run-03_bold.txt');
37. % [mvmt_matrix, mvmt_regs_24, mvmt_table] = movement_regressors(mvmt_file);
38. % % Display the table of movement regressors.
39. % disp(mvmt_table);
40. %
41. % :References:
42. % Power, J.D. et al. (2012). Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion.
43. % Power, J.D. (2019). [Details regarding the correction of respiratory pseudomotion artifacts].
44. %
45. % :See also:
46. % framewise_displacement, outliers, spm_realign, realignment
47. %
49. % More notes:
50. % Power (2012) framewise displacement: Uses the average absolute deviation across 6 movement parameters.
51. % Power 2012: "Spurious but systematic correlations in functional connectivity MRInetworks arisefrom subject motion"
52. % "FDi = |Δdix| + |Δdiy| + |Δdiz| + |Δαi| + |Δβi| + |Δγi|, where Δdix = d(i − 1)x − dix,
53. % and similarly for the other rigid body parameters [dix diy diz αi βi γi].
54. % Rotational displacements were converted from degrees to millimeters by calculating
55. % displace- ment on the surface of a sphere of radius 50 mm, which is approximately
56. % the mean distance from the cerebral cortex to the center of the head." Also: FD
57. % computed with respect to the volume ~2 seconds previous, with respiration pseudo-motion
58. % filtered out of head motion traces (Power 2019).
59. % Power 2019:
60. % "Breathing can cause the head to move, but changes in lung volume can also cause a particular
61. % kind of artifact called pseudomotion, which manifests as a shift of the brain when the lung
62. % expands (Brosch et al., 2002; Durand et al., 2001; Raj et al., 2001)."
63. %
64. % Yoni's protocol for OLP:
65. % spike criteria: FD > .25mm OR abs(zscore(DVARS)) > 3, and the following 4 volumes (following ~2 sec)
66. % bandpass filter, passband = [.01 .1] Hz
68. mvmt_matrix = importdata(mvmt_file);
70. % fill out 24 movement-related parameters per run
71. % ----------------------------------------------------------------
72. mvmt = zscore(mvmt_matrix);
73. mvmt2 = mvmt .^ 2;
74. mvmt3 = [zeros(1, 6); diff(mvmt)];
75. mvmt4 = zscore(mvmt3) .^ 2;
77. % mvmt2 = zscore(mvmt) .^ 2; % Square the Z-scored data
78. % % mvmt3 = [zeros(1, 6); diff(zscore(mvmt))]; % Append zeroes as the first row
79. % mvmt3 = [zeros(1, size(mvmt, 2)); diff(zscore(mvmt))];
80. % mvmt4 = zscore(mvmt3) .^ 2; %
82. mvmt_regs_24 = [zscore(mvmt) mvmt2 mvmt3 mvmt4];
84. names = {'mvmt_x' 'mvmt_y' 'mvmt_z' 'mvmt_roll' 'mvmt_pitch' 'mvmt_yaw' };
85. names = [names 'mvmt_x^2' 'mvmt_y^2' 'mvmt_z^2' 'mvmt_roll^2' 'mvmt_pitch^2' 'mvmt_yaw^2'];
86. names = [names 'mvmt_xdiff' 'mvmt_ydiff' 'mvmt_zdiff' 'mvmt_rolldiff' 'mvmt_pitchdiff' 'mvmt_yawdiff'];
87. names = [names 'mvmt_xdiff^2' 'mvmt_ydiff^2' 'mvmt_zdiff^2' 'mvmt_rolldiff^2' 'mvmt_pitchdiff^2' 'mvmt_yawdiff^2'];
89. % Convert into a table using tasknames as variable names
90. mvmt_table = array2table(mvmt_regs_24, 'VariableNames', names);
92. % % simple geometric mean across translation and rotation; not absolute movement, which varies by voxel
93. % % ----------------------------------------------------------------
94. % geo_displacement = [0; sum(diff(mvmt) .^ 2, 2) .^ .5];
95. %
96. % % adjust outliers: add geo displacement
97. %
98. % high_movement_timepoints(geo_displacement > 0.25) = true;
100. %% New Code
102. % T = size(mvmt, 1); % Number of time points
103. % % len = length(names); % Number of files
104. %
105. % % FDM = zeros(T, len); % Preallocate forward displacement matrix
106. % radius = 50; % Radius for rotation adjustment (in mm)
107. %
108. % % for i = 1:len
109. % % Read each file (assumed to contain motion parameters)
110. % % ts = readmatrix(names{i});
111. %
112. %
113. % % This requires a file with X motion correction parameters from *_LR*.txt
114. % % Select the first 6 columns (translation and rotation) xyz-rpyaw
115. % ts = ts(:, 1:6);
116. %
117. % % Adjust rotation columns (4-6) using the formula for arc length
118. % ts(:, 4:6) = (2 * radius * pi / 360) * ts(:, 4:6);
119. %
120. % % Calculate the difference between consecutive time points
121. % dts = diff(ts);
122. %
123. % % Compute forward displacement as the sum of absolute differences
124. % fwd = sum(abs(dts), 2);
125. %
126. % % Store the forward displacement in FDM, ensuring it fits in the matrix
127. % if length(fwd) == (T - 1)
128. % FDM(2:T, i) = fwd; % Start from the second row, as fwd is one element shorter
129. % end
130. % % end
131. %
132. % % Display the FDM matrix
133. % disp(FDM);
138. end

movement_regressors.m at commit cfc8292, under GPL-2.0 · at the source