OSCR

Mapping the movie-watching brain with AI-derived semantics.

Code ↔ Paper

15 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 15 matches · 2 of them tie a paragraph to a whole file, not to given lines: weak matches, whose lines are not tinted
  1. [1] § Methods › ROI-wise association between AI-encoding performance and cognitive scores ↔ s106_behaviour_motion_control.m, lines 55–114 · score 0.97 · fluid cognition composite, crystalized cognition composite, cognitive flexibility, episodic memory, fluid intelligence, inhibitory control
  2. [2] § Results › AI-derived semantic explainability predicts individual differences in cognitive performance ↔ s106_behaviour_motion_control.m, lines 55–114 · score 0.91 · Fluid Cognition Composite, Crystallized Cognition Composite, Episodic memory, Fluid intelligence, Inhibitory Control, Working memory
  3. [3] § Methods › Movie information and clips generation ↔ s022_gen_clips_overlap.m, lines 70–142 · score 0.88 · sliding window, frame rate, overlap characteristics, fMRI, usable, FFmpeg
  4. [4] § Methods › Clip-wise AI features ↔ Reviewer_2_4.m, lines 27–67 · score 0.84 · scene brightness, motion intensity, social interaction, narrative progress, arousal, music
  5. [5] § Methods › Resting-state FC strength and its association with AI-derived semantic-encoding performance ↔ s105_PLS_strength_r2_motion_control.m, lines 1–35 · score 0.81 · motion corrected residual, nuisance regression, FC strength, head motion, PLS, matrices
  6. [6] § Results › Contribution of semantic feature to the prediction model ↔ Reviewer_2_4.m, lines 27–67 · score 0.79 · Scene brightness, Motion intensity, Social interaction, narrative progress, Arousal, music
  7. [7] § Methods › Clip-wise brain patterns ↔ s026_gen_clippatterns_subject.m, lines 2–26 · score 0.73 · rest frames, fMRI, clip patterns, manifest, trimmed, lagged
  8. [8] § Methods › Clip-wise brain patterns ↔ s023_gen_clippatterns_group.m, lines 2–25 · score 0.73 · rest frames, fMRI, clip patterns, manifest, trimmed, lagged
  9. [9] § Methods › Participants, imaging, and preprocessing ↔ s042_restFC_strength.m, the whole file · a weak match · score 0.72 · linear detrending, 0.01–0.1 Hz, filtering, band, fs, 0.01 Hz
  10. [10] § Methods › ROI-wise semantic encoding using AI features ↔ Reviewer_1_1.m, lines 74–131 · score 0.65 · cross validation, ridge regression, leakage, absolute, fold, training
  11. [11] § Methods › Participants, imaging, and preprocessing ↔ s105_PLS_strength_r2_motion_control.m, lines 1–35 · score 0.64 · nuisance regression, fs LR, motion correction, surface
  12. [12] § Methods › ROI-wise semantic encoding using AI features ↔ s102_regression_group_remove_overlap.m, lines 316–403 · score 0.61 · cross validation, ridge regression, leakage, adjacent, fold, training
  13. [13] § Methods › Resting-state FC strength and its association with AI-derived semantic-encoding performance ↔ s042_restFC_strength.m, the whole file · a weak match · score 0.57 · FC strength, Fisher, transformed, Pearson, matrix, correlations
  14. [14] § Results › Regional variability in semantic predictability across the cortex ↔ Reviewer_1_1.m, lines 74–131 · score 0.53 · Cross validated, ridge regression, features modeled, fit, ROIs, prediction
  15. [15] § Methods › Participants, imaging, and preprocessing ↔ Reviewer_2_5.m, lines 201–256 · score 0.53 · fs LR, workbench, mmp1, CIFTI, hemisphere, parcellation

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

MATLAB · 289 lines · 11 KB · no license · 2 matches

  1. %% =========================================================
  2. % s046_roi_cogcorr_R2_motion_controlled.m
  3. % Method 1: ROI-level Partial Correlation (R² <-> Cognitive Scores)
  4. % Controlling for Head Motion (mean Relative RMS / FD)
  5. % =========================================================
  6. clear; clc; close all;
  7. BaseDir = pwd;
  9. %% ---------- Path Setup ----------
  10. R2Dir = fullfile(BaseDir, '027_roiwise_allclips_subj_no_overlap'); % Upgraded No-Overlap R2
  11. CogCSV = fullfile(BaseDir, 'Subjects_info_7T.csv');
  12. MotionCSV = fullfile(BaseDir, 'subject_mean_motion.csv'); % New Covariate
  13. OutDir = fullfile(BaseDir, '046_roi_cogcorr_R2_motion_controlled');
  15. if ~exist(OutDir, 'dir'), mkdir(OutDir); end
  17. % Cortical surface & parcellation files
  18. Lsurf_file = fullfile(BaseDir, 'fsaverage.L.very_inflated.32k_fs_LR.surf.gii');
  19. Rsurf_file = fullfile(BaseDir, 'fsaverage.R.very_inflated.32k_fs_LR.surf.gii');
  20. wb_cmd = fullfile(BaseDir, '000_workbench','bin_windows64','wb_command.exe');
  21. LabelFile = fullfile(BaseDir, 'HCP_MMP1.0_Glasser.32k_fs_LR.dlabel.nii');
  22. nROI = 360;
  24. %% ====================== 1) Load Motion Data ======================
  25. assert(isfile(MotionCSV), 'Motion CSV not found: %s', MotionCSV);
  26. TMotion = readtable(MotionCSV);
  27. motion_subjID = string(TMotion.subid);
  28. motion_values = TMotion.mean_motion;
  30. % Remove subjects with missing motion data
  31. valid_motion_idx = ~isnan(motion_values);
  32. motion_subjID = motion_subjID(valid_motion_idx);
  33. motion_values = motion_values(valid_motion_idx);
  34. fprintf('Loaded valid head motion data for %d subjects.\n', numel(motion_subjID));
  36. %% ====================== 2) Load R2 (360x1 per subject) ======================
  37. LR2 = dir(fullfile(R2Dir, '*_roiwise_no_overlap.mat'));
  38. assert(~isempty(LR2), 'No *_roiwise_no_overlap.mat found in %s', R2Dir);
  39. R2_subjID = strings(numel(LR2),1);
  40. for i = 1:numel(LR2)
  41. bn = LR2(i).name;
  42. sid = regexp(bn, '^\d+', 'match', 'once');
  43. if isempty(sid)
  44. warning('Cannot extract numeric subject ID from %s, skipping.', bn);
  45. R2_subjID(i) = "";
  46. else
  47. R2_subjID(i) = string(sid);
  48. end
  49. end
  50. validR2 = R2_subjID ~= "";
  51. R2_subjID = R2_subjID(validR2);
  52. LR2 = LR2(validR2);
  53. fprintf('Detected R2 data for %d subjects.\n', numel(R2_subjID));
  55. %% ====================== 3) Load Cognitive Scores CSV ======================
  56. assert(isfile(CogCSV), 'Cognitive scores file not found: %s', CogCSV);
  57. T = readtable(CogCSV);
  58. vnames = T.Properties.VariableNames;
  59. vnames_lower = lower(vnames);
  61. % ---- Automatically find Subject ID column ----
  62. idCand = ["subject","subid","subjectid","sub","id"];
  63. idCol = [];
  64. for k = 1:numel(vnames_lower)
  65. if any(contains(vnames_lower{k}, idCand))
  66. idCol = k;
  67. break;
  68. end
  69. end
  70. assert(~isempty(idCol), 'Subject ID column not detected in Subjects_info_7T.csv.');
  71. subID_cog_raw = T.(vnames{idCol});
  73. % Standardize to string and extract numbers
  74. subID_cog = strings(numel(subID_cog_raw),1);
  75. for i = 1:numel(subID_cog_raw)
  76. x = subID_cog_raw(i);
  77. if isnumeric(x)
  78. subID_cog(i) = string(x);
  79. else
  80. sx = string(x);
  81. tmp = regexp(sx, '\d+', 'match', 'once');
  82. if isempty(tmp)
  83. subID_cog(i) = strtrim(sx);
  84. else
  85. subID_cog(i) = string(tmp);
  86. end
  87. end
  88. end
  90. % ---- Select Cognitive Variables ----
  91. cogVarCands = { ...
  92. 'PMAT24_A_CR', ... % Fluid Intelligence
  93. 'PicSeq_AgeAdj', ... % Episodic Memory
  94. 'ListSort_AgeAdj', ... % Working Memory
  95. 'CardSort_AgeAdj', ... % Cognitive Flexibility
  96. 'Flanker_AgeAdj', ... % Inhibitory Control
  97. 'CogFluidComp_AgeAdj', ... % Fluid Cognition Composite
  98. 'CogCrystalComp_AgeAdj' ... % Crystallized Cognition Composite
  99. };
  100. isCog = ismember(cogVarCands, vnames);
  101. cogVars = cogVarCands(isCog);
  102. Ncog = numel(cogVars);
  103. if isempty(cogVars)
  104. error('Predefined cognitive columns not found in Subjects_info_7T.csv.');
  105. end
  106. fprintf('Using the following %d cognitive variables:\n', Ncog);
  107. disp(cogVars');
  109. cogVars_pretty = strrep(string(cogVars), '_', ' ');
  111. CogMat_all = zeros(height(T), Ncog);
  112. for j = 1:Ncog
  113. CogMat_all(:,j) = T.(cogVars{j});
  114. end
  116. %% ====================== 4) Intersect R2, Cognition, and Motion ======================
  117. [subj_temp, idxR2, idxCog] = intersect(R2_subjID, subID_cog, 'stable');
  118. [subj_common, idx_temp, idxMot] = intersect(subj_temp, motion_subjID, 'stable');
  120. % Remap indices
  121. idxR2 = idxR2(idx_temp);
  122. idxCog = idxCog(idx_temp);
  124. Nsub = numel(subj_common);
  125. assert(Nsub >= 3, 'Not enough overlapping subjects (N=%d).', Nsub);
  126. fprintf('Final intersecting cohort (R2 & Cog & Motion): %d subjects\n', Nsub);
  128. % Assemble R2 Matrix: nROI x Nsub
  129. R2_mat = nan(nROI, Nsub);
  130. for s = 1:Nsub
  131. R2file = fullfile(R2Dir, LR2(idxR2(s)).name);
  132. S_R2 = load(R2file);
  133. if isfield(S_R2, 'R2_s')
  134. r2_vec = S_R2.R2_s;
  135. else
  136. error('R2_s not found in %s', R2file);
  137. end
  138. R2_mat(:, s) = r2_vec(:);
  139. end
  141. % Assemble Cognitive Matrix: Nsub x Ncog
  142. CogMat = CogMat_all(idxCog, :);
  144. % Extract Covariate (Motion): Nsub x 1
  145. Cov_Mot = motion_values(idxMot);
  147. fprintf('R2_mat: %dx%d, CogMat: %dx%d, Motion: %dx1\n', ...
  148. size(R2_mat,1), size(R2_mat,2), size(CogMat,1), size(CogMat,2), size(Cov_Mot,1));
  150. %% ====================== 5) ROI-level PARTIAL Correlation ======================
  151. fprintf('Computing Partial Spearman Correlations (controlling for head motion)...\n');
  152. rho_roi_cog = nan(nROI, Ncog);
  153. p_roi_cog = nan(nROI, Ncog);
  155. for roi = 1:nROI
  156. x = R2_mat(roi, :).'; % Nsub x 1
  157. for j = 1:Ncog
  158. y = CogMat(:, j);
  160. % Filter out NaNs for strict partial correlation
  161. valid_idx = ~isnan(x) & ~isnan(y) & ~isnan(Cov_Mot);
  163. if sum(valid_idx) > 10 % Minimum threshold for statistical validity
  164. [r, p] = partialcorr(x(valid_idx), y(valid_idx), Cov_Mot(valid_idx), 'Type', 'Spearman');
  165. rho_roi_cog(roi, j) = r;
  166. p_roi_cog(roi, j) = p;
  167. end
  168. end
  169. end
  171. %% ====================== 6) FDR Correction (Global across all ROIs & Tests) ======================
  172. p_vec = p_roi_cog(:);
  173. [~, ~, p_adj_vec] = fdr_bh_local(p_vec);
  174. p_roi_cog_FDR = reshape(p_adj_vec, nROI, Ncog);
  175. sig_roi_cog = p_roi_cog_FDR < 0.05;
  177. %% ====================== 7) Export Long Table CSV ======================
  178. ROI_idx_col = repmat((1:nROI).', Ncog, 1);
  179. CogName_col = strings(nROI * Ncog, 1);
  180. rho_col = nan(nROI * Ncog, 1);
  181. p_col = nan(nROI * Ncog, 1);
  182. pFDR_col = nan(nROI * Ncog, 1);
  183. sig_col = false(nROI * Ncog, 1);
  185. cnt = 0;
  186. for j = 1:Ncog
  187. for roi = 1:nROI
  188. cnt = cnt + 1;
  189. CogName_col(cnt) = string(cogVars{j});
  190. rho_col(cnt) = rho_roi_cog(roi, j);
  191. p_col(cnt) = p_roi_cog(roi, j);
  192. pFDR_col(cnt) = p_roi_cog_FDR(roi, j);
  193. sig_col(cnt) = sig_roi_cog(roi, j);
  194. end
  195. end
  196. T_out = table(ROI_idx_col, CogName_col, rho_col, p_col, pFDR_col, sig_col, ...
  197. 'VariableNames', {'ROI','CognitiveMeasure','rho_PartialSpearman','p_raw','p_FDR','Significant_FDRlt0p05'});
  198. writetable(T_out, fullfile(OutDir, 'ROI_cognition_partialcorr_R2.csv'));
  199. fprintf('Exported ROI x Cognition partial correlation results to CSV.\n');
  201. %% ====================== 8) Surface Data Preparation ======================
  202. gL = gifti(Lsurf_file); surf_lh.coord = double(gL.vertices)'; surf_lh.tri = double(gL.faces);
  203. gR = gifti(Rsurf_file); surf_rh.coord = double(gR.vertices)'; surf_rh.tri = double(gR.faces);
  204. L_gii = fullfile(OutDir, 'tmp_MMP.L.32k.label.gii');
  205. R_gii = fullfile(OutDir, 'tmp_MMP.R.32k.label.gii');
  206. cmd = sprintf('"%s" -cifti-separate "%s" COLUMN -label CORTEX_LEFT "%s" -label CORTEX_RIGHT "%s"', ...
  207. wb_cmd, LabelFile, L_gii, R_gii);
  208. system(cmd);
  209. gL = gifti(L_gii); labL = double(gL.cdata(:));
  210. gR = gifti(R_gii); labR = double(gR.cdata(:));
  211. label_32k = [labL; labR];
  212. delete(L_gii); delete(R_gii);
  214. %% ====================== 9) Visualization: Cognitive ROI Maps ======================
  215. if Ncog <= 5
  216. vals_block = rho_roi_cog;
  217. obj1 = plot_hemispheres(vals_block, {surf_lh, surf_rh}, 'parcellation', label_32k, 'colormap', 'jet');
  218. else
  219. nBlock1 = min(5, Ncog);
  220. vals_block1 = rho_roi_cog(:, 1:nBlock1);
  221. obj1 = plot_hemispheres(vals_block1, {surf_lh, surf_rh}, 'parcellation', label_32k, 'colormap', 'jet');
  222. if Ncog > nBlock1
  223. vals_block2 = rho_roi_cog(:, nBlock1+1:end);
  224. obj2 = plot_hemispheres(vals_block2, {surf_lh, surf_rh}, 'parcellation', label_32k, 'colormap', 'jet');
  225. end
  226. end
  228. %% ====================== 10) Extra 1: Significant ROIs per Measure ======================
  229. sig_per_cog = sum(sig_roi_cog, 1);
  230. figure('Color','w','Name','Significant ROIs per cognitive measure (Motion Controlled)');
  231. bar(sig_per_cog, 'FaceColor',[0.3 0.6 0.9],'EdgeColor','none');
  232. set(gca,'XTick',1:Ncog, 'XTickLabel',cogVars_pretty, 'XTickLabelRotation',40, 'FontName','Arial');
  233. ylabel('# significant ROIs (FDR < 0.05)', 'FontName','Arial');
  234. xlabel('Cognitive measure', 'FontName','Arial');
  235. title('Significant ROI counts (Controlling for Head Motion)', 'FontName','Arial');
  236. grid on;
  238. %% ====================== 11) Extra 2: Significant Count per ROI ======================
  239. % sig_count_roi = sum(sig_roi_cog, 2);
  240. % obj_sigcount = plot_hemispheres(sig_count_roi, {surf_lh, surf_rh}, 'parcellation', label_32k);
  241. % title('Number of cognitive scores significantly related to R^2 at each ROI');
  243. %% ====================== 12) Extra 3: Pattern Similarity Matrix ======================
  244. rho_cogpattern = corr(rho_roi_cog, 'rows','pairwise');
  245. figure('Color','w','Name','Similarity between cognitive patterns (Motion Controlled)');
  246. imagesc(rho_cogpattern);
  247. axis square; caxis([-1 1]);
  248. cb = colorbar; cb.Label.String = 'Pattern correlation';
  249. set(gca,'XTick',1:Ncog, 'XTickLabel',cogVars_pretty, 'XTickLabelRotation',40, ...
  250. 'YTick',1:Ncog, 'YTickLabel',cogVars_pretty, 'FontName','Arial');
  251. title('Correlation between ROI-wise patterns (Motion Controlled)', 'FontName','Arial');
  253. %% ====================== 13) Save MAT Results ======================
  254. % save(fullfile(OutDir, 'ROI_cognition_partialcorr_R2.mat'), ...
  255. % 'subj_common', 'R2_mat', 'CogMat', 'Cov_Mot', 'cogVars', 'cogVars_pretty', ...
  256. % 'rho_roi_cog', 'p_roi_cog', 'p_roi_cog_FDR', 'sig_roi_cog', ...
  257. % 'sig_per_cog', 'sig_count_roi', 'rho_cogpattern');
  258. % fprintf('\nCompleted Partial Correlation analysis. Results saved to: %s\n', OutDir);
  260. %% ====================== Local Function: FDR ======================
  261. function [h, crit_p, adj_p] = fdr_bh_local(pvals, q)
  262. if nargin < 2 || isempty(q), q = 0.05; end
  263. p = pvals(:);
  264. valid = ~isnan(p);
  265. p_valid = p(valid);
  266. [p_sorted, sort_ids] = sort(p_valid);
  267. V = length(p_sorted);
  268. I = (1:V)';
  269. cVID = 1;
  270. thresh = I / V * q / cVID;
  271. w = find(p_sorted <= thresh, 1, 'last');
  272. if isempty(w)
  273. crit_p = 0;
  274. h_valid = false(V,1);
  275. else
  276. crit_p = p_sorted(w);
  277. h_valid = p_valid <= crit_p;
  278. end
  279. adj_p_sorted = p_sorted .* V ./ I;
  280. for i = V-1:-1:1
  281. adj_p_sorted(i) = min(adj_p_sorted(i), adj_p_sorted(i+1));
  282. end
  283. adj_p_valid = zeros(V,1);
  284. adj_p_valid(sort_ids) = adj_p_sorted;
  285. h = false(size(p));
  286. adj_p = nan(size(p));
  287. h(valid) = h_valid;
  288. adj_p(valid) = adj_p_valid;
  289. end

s106_behaviour_motion_control.m at commit dd27e15, no license · at the source

Overview

Authors: Muwei Li1,2
ORCID iDs: Muwei Li
  1. Vanderbilt University Institute of Imaging Science, Vanderbilt University Medical Center, Nashville, TN, United States
  2. Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States
Institutions: Vanderbilt University (United States); Vanderbilt University Medical Center (United States)
Journal: Imaging neuroscience (Cambridge, Mass.), volume 4, article IMAG.a.1300
Dates: received 6 December 2025; accepted 22 June 2026; published online 10 July 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1162/imag.a.1300 · PMID 42444710 · PMCID PMC13358718 · OpenAlex W7166184161
Open access: diamond, a free copy (OpenAlex)
Status: code verified
Categories: fMRI (modality), human (organism)
Methods: Spectral & time-frequency, Connectivity, Statistics, Smoothing, state filtering, decompositions, Machine learning, Preprocessing, fMRI & imaging
Keywords: fMRI, naturalistic stimuli, LLMs, semantic encoding
Topic: Functional Brain Connectivity Studies (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: NIMH NIH HHS (U54 MH091657)
Citations: cited by 1 paper (Europe PMC); 64 references in the paper

Abstract

Naturalistic paradigms offer a powerful tool to investigate human brain function, but it remains difficult to link rich, continuous movie content to distributed brain activity in an interpretable way. In this study, I use a multimodal large language model (Gemini) as an automated “semantic annotator” to bridge naturalistic movie stimuli, brain responses, and cognitive performance. Using the Human Connectome Project movie-watching dataset, I segmented the film into 293 overlapping clips, prompting Gemini to rate each clip on 11 psychologically interpretable dimensions. Simultaneously, I extracted clip-wise BOLD activation patterns from the fMR images in 360 cortical ROIs. In this way, the AI and the brain effectively “watch” the same movies in parallel. For each brain ROI, I then fit linear regression models to predict clip-to-clip variation in movie-evoked responses from these features. Gemini-derived features robustly predicted movie-evoked responses in temporal, medial parietal, and lateral frontal association cortex, but explained little variance in unimodal somatosensory, dorsal parietal, insular, and piriform regions. Feature-weight maps reflected known functional specializations, and features with the largest global influence overlapped with the most explainable ROIs. Partial least squares analysis revealed that individual differences in resting-state connectivity strength and semantic explainability covaried along an asymmetric intrinsic axis: strongly integrated sensory-opercular systems at rest were associated with poorer AI predictability, whereas a smaller set of dorsal and medial association regions showed enhanced alignment. Finally, regional AI explainability in medial parietal and left perisylvian association areas was positively related to specific cognitive abilities. Together, these findings demonstrate that interpretable features from AI models provide a simple and scalable framework for quantifying AI-derived semantic predictability in naturalistic settings, offering a practical framework for utilizing artificial models as semantic references to probe human neural processing and individual differences.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repositories

Its files are read in the Code ↔ Paper reader above, with 15 matches between paragraphs and lines of code.

geyerou/Brain-AI-Alignment

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: dd27e154c872d909fc99923426a56ae3c77cebcb, 20 March 2026
Languages: MATLAB (16)
Size: 16 files, 16 scripts
Software Heritage: not archived
Found in: “Data and Code Availability”
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: BrainSpace (9 files), GIfTI library for MATLAB (9 files), Statistics and Machine Learning Toolbox (8 files), Signal Processing Toolbox (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
16 files
At the source: github.com/geyerou/Brain-AI-Alignment

supp:PMC13358718/IMAG.a.1300_supp_Code_SourceFiles_Videos.zip

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Languages: MATLAB (16)
Size: 16 files, 16 scripts
Software Heritage: not checked
Found in: the supplementary material
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: BrainSpace (9 files), GIfTI library for MATLAB (9 files), Statistics and Machine Learning Toolbox (8 files), Signal Processing Toolbox (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)
  • 27 September 2026: the link answers (HTTP 200)
16 files
At the source: pmc-oa-opendata.s3.amazonaws.com/PMC13358718.1/IMAG.a.1300_supp_Code_SourceFiles_Videos.zip

The paper's code and data availability statement is in the Data section.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 2 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 32 scripts, each with its path and the digest of its content;
  • 15 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

Datasets cited

Data and Code Availability

The MRI data used in this study are available in the HCP database https://www.humanconnectome.org/.

To ensure full computational reproducibility, the complete suite of custom MATLAB scripts in this study is provided as a comprehensive supplementary archive accompanying this manuscript. They can also be found at https://github.com/geyerou/Brain-AI-Alignment.

Software and toolboxes that are used in this study:

The Gemini-based feature scoring tool is available upon request from the author. Please provide the email address associated with your own Google Gemini account.

Gemini: https://gemini.google.com/

Google AI Studio: https://aistudio.google.com/

FFmpeg: https://www.ffmpeg.org/

CIFTI: https://www.nitrc.org/projects/cifti/

GIFTI: https://www.nitrc.org/projects/gifti/

HCP workbench: https://www.humanconnectome.org/software/connectome-workbench

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, pages, dates, 1 author, 4 keywords, 1 funder, 64 references.

Cite

This paper

Li, M. (2026). Mapping the movie-watching brain with AI-derived semantics. Imaging neuroscience (Cambridge, Mass.), 4, IMAG.a.1300. https://doi.org/10.1162/imag.a.1300

BibTeX

@article{li2026mapping,
author = {Li, Muwei},
title = {{Mapping the movie-watching brain with AI-derived semantics}},
journal = {Imaging neuroscience (Cambridge, Mass.)},
year = {2026},
month = jul,
volume = {4},
pages = {IMAG.a.1300},
publisher = {MIT Press},
issn = {2837-6056},
doi = {10.1162/imag.a.1300},
url = {https://doi.org/10.1162/imag.a.1300},
pmid = {42444710},
pmcid = {PMC13358718}
}

RIS

TY - JOUR
AU - Li, Muwei
TI - Mapping the movie-watching brain with AI-derived semantics
T2 - Imaging neuroscience (Cambridge, Mass.)
J2 - Imaging Neurosci (Camb)
PY - 2026
DA - 2026/07/10
VL - 4
SP - IMAG.a.1300
SN - 2837-6056
PB - MIT Press
DO - 10.1162/imag.a.1300
UR - https://doi.org/10.1162/imag.a.1300
LA - en
ER -

CSL-JSON

{
"id": "10.1162/imag.a.1300",
"type": "article-journal",
"title": "Mapping the movie-watching brain with AI-derived semantics",
"container-title": "Imaging neuroscience (Cambridge, Mass.)",
"author": [
{
"family": "Li",
"given": "Muwei"
}
],
"container-title-short": "Imaging Neurosci (Camb)",
"volume": "4",
"page": "IMAG.a.1300",
"DOI": "10.1162/imag.a.1300",
"PMID": "42444710",
"PMCID": "PMC13358718",
"ISSN": "2837-6056",
"publisher": "MIT Press",
"URL": "https://doi.org/10.1162/imag.a.1300",
"language": "en",
"issued": {
"date-parts": [
[
2026,
7,
10
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1038/s41467-026-76011-7 [code]
Human cortex organizes dynamic co-fluctuations along the sensorimotor-association axis.
Journal: Nature communications
In common: BrainSpace, GIfTI library for MATLAB, Signal Processing Toolbox, 1 other tool, 4 references
[2] doi:10.1162/imag.a.1331 [code]
Beyond the canonical HRF: Flexible temporal modeling reveals unconstrained BOLD profiles during naturalistic viewing.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: fMRI, 5 references
[3] doi:10.1038/s41467-026-73153-6 [code]
Latent neural architecture organising shared aesthetic evaluations of visual artworks.
Journal: Nature communications
In common: GIfTI library for MATLAB, Statistics and Machine Learning Toolbox, fMRI, 4 references
[4] doi:10.1038/s41467-026-71270-w [code]
Spatiotemporal dynamics of the human cortical functional hierarchy across the lifespan.
Journal: Nature communications
In common: BrainSpace, GIfTI library for MATLAB, Statistics and Machine Learning Toolbox, fMRI, 2 references
[5] doi:10.1002/hbm.70485 [code]
Exploring the Role of the Rich Club in Network Control of Neurocognitive States.
Journal: Human brain mapping
In common: BrainSpace, GIfTI library for MATLAB, Statistics and Machine Learning Toolbox, 2 references
[6] doi:10.1038/s42003-026-10276-y [code]
The cellular correlates and adolescent reorganisation of cortical myelination networks in the common marmoset.
Journal: Communications biology
In common: BrainSpace, GIfTI library for MATLAB, Signal Processing Toolbox, 1 other tool, 1 reference
[7] doi:10.1162/imag.a.1354 [code]
Development of a 24-channel 3T phased-array coil for fMRI in awake monkeys: Mitigating spatiotemporal artifacts in ferumoxytol-weighted functional connectivity estimation.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: fMRI, 6 references
[8] doi:10.1038/s41586-026-10631-3 [code]
A prognostic human brain network for diffuse midline glioma.
Journal: Nature
In common: BrainSpace, GIfTI library for MATLAB, Signal Processing Toolbox, 1 other tool, 1 reference
[9] doi:10.1016/j.neuroimage.2026.122171 [code]
A conserved node degree-based backbone and flexible hub organization of brain connectome during naturalistic movie watching.
Journal: NeuroImage
In common: GIfTI library for MATLAB, Statistics and Machine Learning Toolbox, fMRI, 3 references
[10] doi:10.1038/s41467-026-75959-w [code]
Charting higher-order models of brain function beyond pairwise interactions.
Journal: Nature communications
In common: BrainSpace, Statistics and Machine Learning Toolbox, 4 references

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.