OSCR

Characterizing spatiotemporal white matter hyperintensity pathophysiology in vivo to disentangle vascular and neurodegenerative contributions.

Code ↔ Paper

25 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 25 matches · 5 of them tie a paragraph to a whole file without given lines: weak matches, whose lines are not tinted
  1. [1] § Methods › Associations with vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/5g_pred_ad_stroke.py, lines 336 to 438 · score 0.97 · Lasso logistic regression, class imbalances, fold cross validation, class weights, max_iter, L1 · Report a mismatch
  2. [2] § Results › WMH pathophysiology in vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/5g_pred_ad_stroke.py, lines 336 to 438 · score 0.84 · Lasso logistic regression, fold cross validation, balanced accuracy, F1 score, trained, predictors · Report a mismatch
  3. [3] § Methods › Associations with vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/5b_UKB_disorders.R, lines 43 to 81 · score 0.68 · ischemic heart disease, metabolic disease, lifetime, circulatory, endocrine, nervous · Report a mismatch
  4. [4] § Methods › Spatial clustering ↔ analyses/2_spatial_clust/2j_clust_all_voxels_viz.R, lines 126 to 187 · score 0.67 · variance explained, microstructural metric, ANOVAs, spatial cluster, R2, Voxels · Report a mismatch
  5. [5] § Methods › Spatial clustering ↔ analyses/3_temporal_sustain/3f_clean_wmh_data.R, the whole file · a weak match · score 0.67 · log transformed, common space, spatial cluster, WMH volumes, WMH pathophysiology, median · Report a mismatch
  6. [6] § Methods › Associations with vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/5a_describe_UKB_ICD10_categ.R, lines 56 to 102 · score 0.67 · ischemic heart disease, metabolic disease, lifetime, circulatory, endocrine, nervous · Report a mismatch
  7. [7] § Results › Estimates of WMH pathophysiology ↔ analyses/3_temporal_sustain/3e_WMHpatho_with_volume.R, lines 54 to 117 · score 0.66 · fluid sensitive, iron sensitive, biological sensitivity, fiber sensitive, myelin, scored · Report a mismatch
  8. [8] § Results › Defining temporal pathophysiological cascades ↔ analyses/3_temporal_sustain/3d_sustain_viz.R, lines 293 to 334 · score 0.66 · Positional variance diagrams, cross validated, curves, SuStaIn, temporal, QSM · Report a mismatch
  9. [9] § Methods › Temporal modeling ↔ analyses/3_temporal_sustain/3f_clean_wmh_data.R, the whole file · a weak match · score 0.64 · log transformed WMH, SuStaIn, WMH volume, WMH pathophysiological, temporal, cluster · Report a mismatch
  10. [10] § Methods › Spatial clustering ↔ analyses/2_spatial_clust/2d_assign_excluded_voxels.R, lines 2 to 61 · score 0.64 · NAWM prevalence, excluded voxel, UKB template, surrounding, iteratively, clusters · Report a mismatch
  11. [11] § Results › Spatial regions of pathophysiologically similar WMHs ↔ analyses/2_spatial_clust/2e_overlap_other_parc.R, lines 2 to 47 · score 0.64 · driven cluster, commissural, parietal, frontal, quantify, lobar · Report a mismatch
  12. [12] § Results › Image processing and microstructural MRI ↔ analyses/3_temporal_sustain/3e_WMHpatho_with_volume.R, lines 54 to 117 · score 0.63 · fluid sensitive, iron sensitive, fiber sensitive, myelin, biological, volume · Report a mismatch
  13. [13] § Results › Posterior WMHs connect cortical regions prone to tau pathology ↔ analyses/6_grey_matter/6b_connected_gm_viz.R, the whole file · a weak match · score 0.63 · Pearson correlations, Bonferroni corrected, Tau PET, connecting, amyloid, Posterior · Report a mismatch
  14. [14] § Methods › Associations with vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/5e_correlate_patterns.R, lines 117 to 179 · score 0.61 · correlation matrix, Hierarchical clustering, hclust, Pearson, ADNI, disorders · Report a mismatch
  15. [15] § Methods › Associations with vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/parcnofill_correlate_patterns.R, lines 118 to 180 · score 0.61 · correlation matrix, Hierarchical clustering, hclust, Pearson, ADNI, disorders · Report a mismatch
  16. [16] § Results › Spatial regions of pathophysiologically similar WMHs ↔ analyses/2_spatial_clust/2d_assign_excluded_voxels.R, lines 2 to 61 · score 0.57 · closest spatial cluster, WMH prevalence, dimensionality, voxels · Report a mismatch
  17. [17] § Methods › Associations with vascular and neurodegenerative disorders ↔ analyses/3_temporal_sustain/3c_sustain_bind_CV.py, lines 112 to 125 · score 0.56 · cross validation, scikit-learn, sklearn, train, fold, models · Report a mismatch
  18. [18] § Results › WMH pathophysiology in vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/5h_pred_ad_stroke_viz.R, lines 49 to 105 · score 0.56 · balanced accuracy, AUROC, Wilcoxon, Bonferroni, SD, fold · Report a mismatch
  19. [19] § Methods › Spatial clustering ↔ analyses/2_spatial_clust/2h_roi_avg.R, the whole file · a weak match · score 0.56 · common space, WMH voxels, WMH volumes, WMH pathophysiology, median, ICVF · Report a mismatch
  20. [20] § Results › Spatial regions of pathophysiologically similar WMHs ↔ analyses/2_spatial_clust/2f_clust_by_WMH_volume.R, lines 167 to 206 · score 0.53 · spectral clustering, cluster solution, severity, NAWM, prevalence, QSM · Report a mismatch
  21. [21] § Results › WMH pathophysiology in vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/5e_correlate_patterns.R, lines 2 to 73 · score 0.52 · ischemic heart disease, cardiovascular disease, correlated, Alzheimer, PRS, stroke · Report a mismatch
  22. [22] § Results › WMH pathophysiology in vascular and neurodegenerative disorders ↔ analyses/5_WMH_across_disorders/parcnofill_correlate_patterns.R, lines 2 to 74 · score 0.52 · ischemic heart disease, cardiovascular disease, correlated, Alzheimer, PRS, stroke · Report a mismatch
  23. [23] § Methods › Image processing ↔ analyses/1_norm_mod/1d_zscore_ADNI.R, lines 2 to 58 · score 0.52 · UKB space, WMH masks, BISON, tIssue, microstructural, ADNI · Report a mismatch
  24. [24] § Results › Spatial regions of pathophysiologically similar WMHs ↔ analyses/2_spatial_clust/2i_spectral_clust_all_voxels.R, the whole file · a weak match · score 0.52 · spectral clustering, WMH pathophysiology maps, NAWM, prevalence, QSM, ISOVF · Report a mismatch
  25. [25] § Methods › Temporal modeling ↔ analyses/3_temporal_sustain/3b_sustain_run.py, lines 62 to 125 · score 0.50 · pySuStaIn, startpoints, fold, temporal, model, WMH · Report a mismatch

Paper

Loaded from Europe PMC by your browser; OSCR does not store it: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Python · 480 lines · 20 KB · no license · 2 matches

  2. import numpy as np
  3. import pandas as pd
  4. import matplotlib.pyplot as plt
  5. import seaborn as sns
  6. from sklearn.model_selection import StratifiedKFold, cross_validate, train_test_split
  7. from sklearn.linear_model import LogisticRegression
  8. from sklearn.preprocessing import StandardScaler
  9. from sklearn.metrics import (
  10. classification_report, confusion_matrix, roc_auc_score,
  11. roc_curve, precision_recall_curve, average_precision_score,
  12. make_scorer, f1_score
  13. )
  14. from sklearn.utils.class_weight import compute_class_weight
  15. import warnings
  16. import datatable as dt
  17. from neuroHarmonize import harmonizationLearn, harmonizationApply
  18. from sklearn.metrics import accuracy_score, precision_score, recall_score, balanced_accuracy_score
  20. warnings.filterwarnings('ignore')
  22. # Set random seed for reproducibility
  23. np.random.seed(42)
  25. #region Load data for spect_clust_k3 and ComBat harmonization with ADNI
  27. # Load UKB stroke data
  28. df_ukb = pd.read_csv(f"../3_temporal_sustain/results/3f_clean_wmh_data/wmh_combined.tsv", delimiter="\t")
  29. df_ukb = df_ukb.drop(columns=['stage_c1', 'stage_c2', 'stage_c3', 'post_ant_subtype', 'post_ant_stage', 'T2star_c1', 'T2star_c2', 'T2star_c3', 'QSM_c1', 'QSM_c2', 'QSM_c3'])
  31. df_dx = dt.fread("../../../UKB/Analyses/clean_firstocc/results/firstocc_categ.tsv").to_pandas()
  32. df_dx = df_dx[df_dx["chapter"].isin([
  33. "Endocrine, nutritional and metabolic diseases",
  34. "Mental, Behavioral and Neurodevelopmental disorders",
  35. "Diseases of the circulatory system",
  36. "Diseases of the nervous system"
  37. ])]
  38. df_dx = df_dx.merge(df_ukb['ID'], how="right")
  39. controls = df_dx[df_dx.isna().any(axis=1)]["ID"].unique()
  40. df_ukb_combat = df_ukb[df_ukb["ID"].isin(controls)].reset_index(drop=True)
  42. stroke_cases = (df_dx.loc[df_dx["icd_code"].isin(["I60", "I61", "I63", "I64"]), ["ID"]].drop_duplicates().reset_index(drop=True))
  43. dementia_cases = (df_dx.loc[df_dx["icd_code"].isin(["F00","F03","G30"]), ["ID"]].drop_duplicates().reset_index(drop=True))
  45. overlap_ids = set(stroke_cases['ID']).intersection(dementia_cases['ID'])
  46. stroke_cases = stroke_cases[~stroke_cases['ID'].isin(overlap_ids)].reset_index(drop=True)
  47. dementia_cases = dementia_cases[~dementia_cases['ID'].isin(overlap_ids)].reset_index(drop=True)
  49. df_stroke = pd.merge(stroke_cases, df_ukb)
  50. df_stroke = df_stroke.drop(columns=['ID'])
  51. df_stroke['Group'] = 'Stroke'
  52. df_stroke['SITE'] = 'UKB'
  54. df_dementia_ukb = pd.merge(dementia_cases, df_ukb)
  55. df_dementia_ukb = df_dementia_ukb.drop(columns=['ID'])
  56. df_dementia_ukb['Group'] = 'cog_imp'
  57. df_dementia_ukb['SITE'] = 'UKB'
  59. # Load ADNI data
  60. df_adni = pd.read_csv(f"../clean_wmh_data/results/spect_clust_k3/wmh_combined_ADNI.tsv", delimiter="\t")
  61. df_adni = df_adni.rename(columns={"Age": "age", "Sex": "sex"})
  63. df_demo = pd.read_csv("../../../ADNI/Analyses/clean_data/results/df_demo.tsv", delimiter="\t")
  64. df_demo = df_demo.rename(columns={"PTID": "ID", "EXAMDATE": "date"})
  65. # Keep last timepoint only
  66. df_demo = (
  67. df_demo
  68. .sort_values(['ID']) # optional unless order matters within ID
  69. .groupby('ID', as_index=False)
  70. .tail(1)
  71. .reset_index(drop=True)
  72. )
  73. # Remove CI due to non AD
  74. df_demo = df_demo[
  75. (df_demo['dx_mci_cause'].isna() | (df_demo['dx_mci_cause'] != "Due to other")) &
  76. (df_demo['dx_dementia_cause'].isna() | (df_demo['dx_dementia_cause'] != "Due to other"))
  77. ]
  78. # Merge with WMH data
  79. df_demo_merge = df_demo.loc[:,['ID', 'date']]
  80. df_adni = pd.merge(df_demo_merge, df_adni, on=['ID', 'date'])
  81. df_adni = df_adni.dropna().reset_index(drop=True)
  82. df_demo = pd.merge(df_demo, df_adni.loc[:,['ID', 'date']], on=['ID', 'date'])
  84. # Keep only MCI or dementia
  85. df_demo_merge = df_demo[df_demo['dx'].isin(['MCI', 'Dementia'])].reset_index(drop=True)
  86. df_demo_merge = df_demo_merge.loc[:,["ID", "date"]]
  88. df_ci_adni = pd.merge(df_demo_merge, df_adni, how="left", on=['ID', 'date'])
  89. df_ci_adni['Group'] = 'cog_imp'
  90. df_ci_adni['SITE'] = 'ADNI'
  92. df_ci_adni = df_ci_adni.loc[:,df_stroke.columns]
  94. # Select healthy IDs from ADNI
  95. hc_ids = df_demo[(df_demo["dx"] == "CN")]['ID']
  96. df_adni_combat = df_adni[df_adni["ID"].isin(hc_ids)].reset_index(drop=True)
  97. df_adni_combat = df_adni_combat.drop(columns=["date", "Year"])
  99. # ComBat harmonization
  100. df_adni_combat['SITE'] = 'ADNI'
  101. df_ukb_combat['SITE'] = 'UKB'
  102. df_combat = pd.concat([df_adni_combat, df_ukb_combat], ignore_index=True)
  103. df_combat['sex'] = df_combat['sex'].map({'Male': 1, 'Female': 0}).values
  105. exclude_cols = ["ID", "age", "sex", "SITE"]
  106. feature_cols = [c for c in df_combat.columns if c not in exclude_cols]
  107. X_combat = df_combat[feature_cols].to_numpy()
  108. covars = df_combat.loc[:,['SITE', 'age', 'sex']]
  110. # Train
  111. combat_model, df_combat_adj = harmonizationLearn(X_combat, covars, ref_batch="UKB")
  113. # Predict
  114. df_all = pd.concat([df_stroke, df_dementia_ukb, df_ci_adni], axis=0, ignore_index=True)
  115. df_all['sex'] = df_all['sex'].map({'Male': 1, 'Female': 0}).values
  117. exclude_cols = ["age", "sex", 'Group', "SITE"]
  118. feature_cols = [c for c in df_all.columns if c not in exclude_cols]
  119. X_combat = df_all[feature_cols].to_numpy()
  120. covars = df_all.loc[:,['SITE', 'age', 'sex']]
  122. X_clust = harmonizationApply(X_combat, covars, combat_model)
  123. X_clust_vol = X_clust[:,0:3]
  124. y_clust = df_all['Group'].map({'Stroke': 0, 'cog_imp': 1}).values
  126. #endregion
  128. #region Load data for PV/deep and ComBat harmonization with ADNI
  130. # Load UKB stroke data
  131. df_ukb = pd.read_csv(f"../clean_wmh_data/results/pv_deep/wmh_combined.tsv", delimiter="\t")
  132. df_ukb = df_ukb.drop(columns=['pv_deep_T2star_c1', 'pv_deep_T2star_c2', 'pv_deep_QSM_c1', 'pv_deep_QSM_c2'])
  134. df_dx = dt.fread("../../../UKB/Analyses/clean_firstocc/results/firstocc_categ.tsv").to_pandas()
  135. df_dx = df_dx[df_dx["chapter"].isin([
  136. "Endocrine, nutritional and metabolic diseases",
  137. "Mental, Behavioral and Neurodevelopmental disorders",
  138. "Diseases of the circulatory system",
  139. "Diseases of the nervous system"
  140. ])]
  141. df_dx = df_dx.merge(df_ukb['ID'], how="right")
  142. controls = df_dx[df_dx.isna().any(axis=1)]["ID"].unique()
  143. df_ukb_combat = df_ukb[df_ukb["ID"].isin(controls)].reset_index(drop=True)
  145. stroke_cases = (df_dx.loc[df_dx["icd_code"].isin(["I60", "I61", "I63", "I64"]), ["ID"]].drop_duplicates().reset_index(drop=True))
  146. dementia_cases = (df_dx.loc[df_dx["icd_code"].isin(["F00","F03","G30"]), ["ID"]].drop_duplicates().reset_index(drop=True))
  148. overlap_ids = set(stroke_cases['ID']).intersection(dementia_cases['ID'])
  149. stroke_cases = stroke_cases[~stroke_cases['ID'].isin(overlap_ids)].reset_index(drop=True)
  150. dementia_cases = dementia_cases[~dementia_cases['ID'].isin(overlap_ids)].reset_index(drop=True)
  152. df_stroke = pd.merge(stroke_cases, df_ukb)
  153. df_stroke = df_stroke.drop(columns=['ID'])
  154. df_stroke['Group'] = 'Stroke'
  155. df_stroke['SITE'] = 'UKB'
  157. df_dementia_ukb = pd.merge(dementia_cases, df_ukb)
  158. df_dementia_ukb = df_dementia_ukb.drop(columns=['ID'])
  159. df_dementia_ukb['Group'] = 'cog_imp'
  160. df_dementia_ukb['SITE'] = 'UKB'
  162. # Load ADNI data
  163. df_adni = pd.read_csv(f"../clean_wmh_data/results/pv_deep/wmh_combined_ADNI.tsv", delimiter="\t")
  164. df_adni = df_adni.rename(columns={"Age": "age", "Sex": "sex"})
  166. df_demo = pd.read_csv("../../../ADNI/Analyses/clean_data/results/df_demo.tsv", delimiter="\t")
  167. df_demo = df_demo.rename(columns={"PTID": "ID", "EXAMDATE": "date"})
  168. # Keep last timepoint only
  169. df_demo = (
  170. df_demo
  171. .sort_values(['ID']) # optional unless order matters within ID
  172. .groupby('ID', as_index=False)
  173. .tail(1)
  174. .reset_index(drop=True)
  175. )
  176. # Remove CI due to non AD
  177. df_demo = df_demo[
  178. (df_demo['dx_mci_cause'].isna() | (df_demo['dx_mci_cause'] != "Due to other")) &
  179. (df_demo['dx_dementia_cause'].isna() | (df_demo['dx_dementia_cause'] != "Due to other"))
  180. ]
  181. # Merge with WMH data
  182. df_demo_merge = df_demo.loc[:,['ID', 'date']]
  183. df_adni = pd.merge(df_demo_merge, df_adni, on=['ID', 'date'])
  184. df_adni = df_adni.dropna().reset_index(drop=True)
  185. df_demo = pd.merge(df_demo, df_adni.loc[:,['ID', 'date']], on=['ID', 'date'])
  187. # Keep only MCI or dementia
  188. df_demo_merge = df_demo[df_demo['dx'].isin(['MCI', 'Dementia'])].reset_index(drop=True)
  189. df_demo_merge = df_demo_merge.loc[:,["ID", "date"]]
  191. df_ci_adni = pd.merge(df_demo_merge, df_adni, how="left", on=['ID', 'date'])
  192. df_ci_adni['Group'] = 'cog_imp'
  193. df_ci_adni['SITE'] = 'ADNI'
  195. df_ci_adni = df_ci_adni.loc[:,df_stroke.columns]
  197. # Select healthy IDs from ADNI
  198. hc_ids = df_demo[(df_demo["dx"] == "CN")]['ID']
  199. df_adni_combat = df_adni[df_adni["ID"].isin(hc_ids)].reset_index(drop=True)
  200. df_adni_combat = df_adni_combat.drop(columns=["date", "Year"])
  202. # ComBat harmonization
  203. df_adni_combat['SITE'] = 'ADNI'
  204. df_ukb_combat['SITE'] = 'UKB'
  205. df_combat = pd.concat([df_adni_combat, df_ukb_combat], ignore_index=True)
  206. df_combat['sex'] = df_combat['sex'].map({'Male': 1, 'Female': 0}).values
  208. exclude_cols = ["ID", "age", "sex", "SITE"]
  209. feature_cols = [c for c in df_combat.columns if c not in exclude_cols]
  210. X_combat = df_combat[feature_cols].to_numpy()
  211. covars = df_combat.loc[:,['SITE', 'age', 'sex']]
  213. # Train
  214. combat_model, df_combat_adj = harmonizationLearn(X_combat, covars, ref_batch="UKB")
  216. # Predict
  217. df_all = pd.concat([df_stroke, df_dementia_ukb, df_ci_adni], axis=0, ignore_index=True)
  218. df_all['sex'] = df_all['sex'].map({'Male': 1, 'Female': 0}).values
  220. exclude_cols = ["age", "sex", 'Group', "SITE"]
  221. feature_cols = [c for c in df_all.columns if c not in exclude_cols]
  222. X_combat = df_all[feature_cols].to_numpy()
  223. covars = df_all.loc[:,['SITE', 'age', 'sex']]
  225. X_pvdeep = harmonizationApply(X_combat, covars, combat_model)
  226. X_pvdeep_vol = X_pvdeep[:,0:2]
  227. y_pvdeep = df_all['Group'].map({'Stroke': 0, 'cog_imp': 1}).values
  229. #endregion
  231. #region Load data for no parcellation and ComBat harmonization with ADNI
  233. # Load UKB stroke data
  234. df_ukb = pd.read_csv(f"../clean_wmh_data/results/no_clust/wmh_combined.tsv", delimiter="\t")
  235. df_ukb = df_ukb.drop(columns=['stage_c1', 'T2star_c1', 'QSM_c1'])
  237. df_dx = dt.fread("../../../UKB/Analyses/clean_firstocc/results/firstocc_categ.tsv").to_pandas()
  238. df_dx = df_dx[df_dx["chapter"].isin([
  239. "Endocrine, nutritional and metabolic diseases",
  240. "Mental, Behavioral and Neurodevelopmental disorders",
  241. "Diseases of the circulatory system",
  242. "Diseases of the nervous system"
  243. ])]
  244. df_dx = df_dx.merge(df_ukb['ID'], how="right")
  245. controls = df_dx[df_dx.isna().any(axis=1)]["ID"].unique()
  246. df_ukb_combat = df_ukb[df_ukb["ID"].isin(controls)].reset_index(drop=True)
  248. stroke_cases = (df_dx.loc[df_dx["icd_code"].isin(["I60", "I61", "I63", "I64"]), ["ID"]].drop_duplicates().reset_index(drop=True))
  249. dementia_cases = (df_dx.loc[df_dx["icd_code"].isin(["F00","F03","G30"]), ["ID"]].drop_duplicates().reset_index(drop=True))
  251. overlap_ids = set(stroke_cases['ID']).intersection(dementia_cases['ID'])
  252. stroke_cases = stroke_cases[~stroke_cases['ID'].isin(overlap_ids)].reset_index(drop=True)
  253. dementia_cases = dementia_cases[~dementia_cases['ID'].isin(overlap_ids)].reset_index(drop=True)
  255. df_stroke = pd.merge(stroke_cases, df_ukb)
  256. df_stroke = df_stroke.drop(columns=['ID'])
  257. df_stroke['Group'] = 'Stroke'
  258. df_stroke['SITE'] = 'UKB'
  260. df_dementia_ukb = pd.merge(dementia_cases, df_ukb)
  261. df_dementia_ukb = df_dementia_ukb.drop(columns=['ID'])
  262. df_dementia_ukb['Group'] = 'cog_imp'
  263. df_dementia_ukb['SITE'] = 'UKB'
  265. # Load ADNI data
  266. df_adni = pd.read_csv(f"../clean_wmh_data/results/no_clust/wmh_combined_ADNI.tsv", delimiter="\t")
  267. df_adni = df_adni.rename(columns={"Age": "age", "Sex": "sex", "region_vol[, seq(2, ncol(region_vol))]": "WMHvol_c1"})
  269. df_demo = pd.read_csv("../../../ADNI/Analyses/clean_data/results/df_demo.tsv", delimiter="\t")
  270. df_demo = df_demo.rename(columns={"PTID": "ID", "EXAMDATE": "date"})
  271. # Keep last timepoint only
  272. df_demo = (
  273. df_demo
  274. .sort_values(['ID']) # optional unless order matters within ID
  275. .groupby('ID', as_index=False)
  276. .tail(1)
  277. .reset_index(drop=True)
  278. )
  279. # Remove CI due to non AD
  280. df_demo = df_demo[
  281. (df_demo['dx_mci_cause'].isna() | (df_demo['dx_mci_cause'] != "Due to other")) &
  282. (df_demo['dx_dementia_cause'].isna() | (df_demo['dx_dementia_cause'] != "Due to other"))
  283. ]
  284. # Merge with WMH data
  285. df_demo_merge = df_demo.loc[:,['ID', 'date']]
  286. df_adni = pd.merge(df_demo_merge, df_adni, on=['ID', 'date'])
  287. df_adni = df_adni.dropna().reset_index(drop=True)
  288. df_demo = pd.merge(df_demo, df_adni.loc[:,['ID', 'date']], on=['ID', 'date'])
  290. # Keep only MCI or dementia
  291. df_demo_merge = df_demo[df_demo['dx'].isin(['MCI', 'Dementia'])].reset_index(drop=True)
  292. df_demo_merge = df_demo_merge.loc[:,["ID", "date"]]
  294. df_ci_adni = pd.merge(df_demo_merge, df_adni, how="left", on=['ID', 'date'])
  295. df_ci_adni['Group'] = 'cog_imp'
  296. df_ci_adni['SITE'] = 'ADNI'
  298. df_ci_adni = df_ci_adni.loc[:,df_stroke.columns]
  300. # Select healthy IDs from ADNI
  301. hc_ids = df_demo[(df_demo["dx"] == "CN")]['ID']
  302. df_adni_combat = df_adni[df_adni["ID"].isin(hc_ids)].reset_index(drop=True)
  303. df_adni_combat = df_adni_combat.drop(columns=["date", "Year"])
  305. # ComBat harmonization
  306. df_adni_combat['SITE'] = 'ADNI'
  307. df_ukb_combat['SITE'] = 'UKB'
  308. df_combat = pd.concat([df_adni_combat, df_ukb_combat], ignore_index=True)
  309. df_combat['sex'] = df_combat['sex'].map({'Male': 1, 'Female': 0}).values
  311. exclude_cols = ["ID", "age", "sex", "SITE"]
  312. feature_cols = [c for c in df_combat.columns if c not in exclude_cols]
  313. X_combat = df_combat[feature_cols].to_numpy()
  314. covars = df_combat.loc[:,['SITE', 'age', 'sex']]
  316. # Train
  317. combat_model, df_combat_adj = harmonizationLearn(X_combat, covars, ref_batch="UKB")
  319. # Predict
  320. df_all = pd.concat([df_stroke, df_dementia_ukb, df_ci_adni], axis=0, ignore_index=True)
  321. df_all['sex'] = df_all['sex'].map({'Male': 1, 'Female': 0}).values
  323. exclude_cols = ["age", "sex", 'Group', "SITE"]
  324. feature_cols = [c for c in df_all.columns if c not in exclude_cols]
  325. X_combat = df_all[feature_cols].to_numpy()
  326. covars = df_all.loc[:,['SITE', 'age', 'sex']]
  328. X_noparc = harmonizationApply(X_combat, covars, combat_model)
  329. X_noparc_vol = X_noparc[:,0]
  330. y_noparc = df_all['Group'].map({'Stroke': 0, 'cog_imp': 1}).values
  332. #endregion
  334. #region Function to do model fit and CV
  336. def fit_repeated_cv(X, y, n_repeats=100, cv=5):
  337. """
  338. Perform repeated stratified k-fold cross-validation
  340. Parameters:
  341. -----------
  342. X : numpy array
  343. Feature matrix
  344. y : numpy array
  345. Target vector
  346. n_repeats : int
  347. Number of times to repeat cross-validation
  348. cv : int
  349. Number of folds per repeat
  350. model_type : str
  351. 'lasso' for Lasso Logistic Regression or 'rf' for Random Forest
  352. """
  353. from sklearn.metrics import accuracy_score, precision_score, recall_score, balanced_accuracy_score
  354. from sklearn.ensemble import RandomForestClassifier
  355. # Store metrics from all repeats
  356. all_repeat_metrics = {
  357. 'balanced_accuracy': [],
  358. 'f1': [], 'roc_auc': []
  359. }
  360. # Store predictions from first repeat only (for visualization)
  361. first_repeat_y_true = []
  362. first_repeat_y_pred = []
  363. first_repeat_y_pred_proba = []
  364. print("\n" + "="*60)
  365. model_name = "LASSO LOGISTIC REGRESSION"
  366. print(f"REPEATED CROSS-VALIDATION - {model_name}")
  367. print(f"({n_repeats} x {cv}-Fold Stratified)")
  368. print("="*60)
  369. print(f"Total model fits: {n_repeats * cv} = {n_repeats * cv}")
  370. # Compute class weights
  371. class_weights = compute_class_weight('balanced',
  372. classes=np.unique(y),
  373. y=y)
  374. class_weight_dict = {0: class_weights[0], 1: class_weights[1]}
  375. print(f"Class weights for imbalance: {class_weight_dict}")
  376. for repeat in range(n_repeats):
  377. # Use different random state for each repeat
  378. skf = StratifiedKFold(n_splits=cv, shuffle=True, random_state=repeat)
  379. repeat_metrics = {
  380. 'balanced_accuracy': [],
  381. 'accuracy': [], 'precision': [], 'recall': [],
  382. 'f1': [], 'roc_auc': []
  383. }
  384. for fold, (train_idx, test_idx) in enumerate(skf.split(X, y), 1):
  385. # Split data
  386. X_train_fold, X_test_fold = X[train_idx], X[test_idx]
  387. y_train_fold, y_test_fold = y[train_idx], y[test_idx]
  388. # Standardize features (fit on train, transform both)
  389. scaler = StandardScaler()
  390. X_train_scaled = scaler.fit_transform(X_train_fold)
  391. X_test_scaled = scaler.transform(X_test_fold)
  392. # Train model
  393. model = LogisticRegression(
  394. penalty='l1',
  395. C=1.0,
  396. solver='saga',
  397. class_weight='balanced', # HANDLES CLASS IMBALANCE
  398. max_iter=1000,
  399. random_state=42,
  400. n_jobs=-1
  401. )
  402. model.fit(X_train_scaled, y_train_fold)
  403. # Predict
  404. y_pred = model.predict(X_test_scaled)
  405. y_pred_proba = model.predict_proba(X_test_scaled)[:, 1]
  406. # Store predictions from first repeat only
  407. if repeat == 0:
  408. first_repeat_y_true.extend(y_test_fold)
  409. first_repeat_y_pred.extend(y_pred)
  410. first_repeat_y_pred_proba.extend(y_pred_proba)
  411. # Calculate metrics for this fold
  412. repeat_metrics['balanced_accuracy'].append(balanced_accuracy_score(y_test_fold, y_pred))
  413. repeat_metrics['f1'].append(f1_score(y_test_fold, y_pred, zero_division=0))
  414. repeat_metrics['roc_auc'].append(roc_auc_score(y_test_fold, y_pred_proba))
  415. # Store mean metrics for this repeat
  416. for metric in repeat_metrics:
  417. all_repeat_metrics[metric].append(np.mean(repeat_metrics[metric]))
  418. # Progress update every 10 repeats
  419. if (repeat + 1) % 10 == 0:
  420. current_auc = np.mean(all_repeat_metrics['roc_auc'])
  421. print(f" Completed {repeat + 1}/{n_repeats} repeats... "
  422. f"Mean ROC-AUC: {current_auc:.3f}")
  423. # Convert first repeat predictions to arrays
  424. first_repeat_y_true = np.array(first_repeat_y_true)
  425. first_repeat_y_pred = np.array(first_repeat_y_pred)
  426. first_repeat_y_pred_proba = np.array(first_repeat_y_pred_proba)
  427. # Print summary statistics across all repeats
  428. print("\n" + "-"*60)
  429. print(f"SUMMARY ACROSS {n_repeats} REPEATS")
  430. print("-"*60)
  431. for metric in ['balanced_accuracy', 'f1', 'roc_auc']:
  432. scores = all_repeat_metrics[metric]
  433. print(f"{metric.upper()}:")
  434. print(f" Mean: {np.mean(scores):.4f}")
  435. print(f" Std: {np.std(scores):.4f}")
  436. print(f" 95% CI: [{np.percentile(scores, 2.5):.4f}, {np.percentile(scores, 97.5):.4f}]")
  437. return (first_repeat_y_true, first_repeat_y_pred, first_repeat_y_pred_proba,
  438. all_repeat_metrics)
  440. #endregion
  442. #region Run
  444. ################# Spect_clust_k3
  446. # Volumes + micro
  447. y_true, y_pred, y_pred_proba, fold_metrics = fit_repeated_cv(X_clust, y_clust, cv=5, model_type='lasso')
  448. df_metrics = pd.DataFrame(fold_metrics)
  449. df_metrics.to_csv("./results/5g_pred_ad_stroke/spect_clust_k3_micro_metrics.csv", index=False)
  451. # Volumes only
  452. y_true, y_pred, y_pred_proba, fold_metrics = fit_repeated_cv(X_clust_vol, y_clust, cv=5, model_type='lasso')
  453. df_metrics = pd.DataFrame(fold_metrics)
  454. df_metrics.to_csv("./results/5g_pred_ad_stroke/spect_clust_k3_vol_metrics.csv", index=False)
  456. ################# PV/deep
  458. # Volumes + micro
  459. y_true, y_pred, y_pred_proba, fold_metrics = fit_repeated_cv(X_pvdeep, y_pvdeep, cv=5, model_type='lasso')
  460. df_metrics = pd.DataFrame(fold_metrics)
  461. df_metrics.to_csv("./results/5g_pred_ad_stroke/pv_deep_micro_metrics.csv", index=False)
  463. # Volumes only
  464. y_true, y_pred, y_pred_proba, fold_metrics = fit_repeated_cv(X_pvdeep_vol, y_pvdeep, cv=5, model_type='lasso')
  465. df_metrics = pd.DataFrame(fold_metrics)
  466. df_metrics.to_csv("./results/5g_pred_ad_stroke/pv_deep_vol_metrics.csv", index=False)
  468. ################# No parcellation
  470. # Volumes + micro
  471. y_true, y_pred, y_pred_proba, fold_metrics = fit_repeated_cv(X_noparc, y_noparc, cv=5, model_type='lasso')
  472. df_metrics = pd.DataFrame(fold_metrics)
  473. df_metrics.to_csv("./results/5g_pred_ad_stroke/noparc_micro_metrics.csv", index=False)
  475. # Volumes only
  476. y_true, y_pred, y_pred_proba, fold_metrics = fit_repeated_cv(X_noparc_vol.reshape(-1,1), y_noparc, cv=5)
  477. df_metrics = pd.DataFrame(fold_metrics)
  478. df_metrics.to_csv("./results/5g_pred_ad_stroke/noparc_vol_metrics.csv", index=False)
  480. #endregion

5g_pred_ad_stroke.py at commit ad83218, no license · at the source

Overview

Authors: Olivier Parent1,2, Zaki Alasmar1,2, Sophia Osborne1, Aurélie Bussy1,2, Manuela Costantino1, Jérémie P Fouquet1, Daniela Quesada2, Alexandre Pastor-Bernier3,4, Alfonso Fajardo-Valdez3, Alexa Pichet-Binette5,6, Ann McQuarrie7, Josefina Maranzano8,9, Gabriel A Devenyi1,10, Christopher J Steele11,12, Sylvia Villeneuve3,10,13, the PREVENT-AD Research Group, the Alzheimer’s Disease Neuroimaging Initiative (ADNI), Mahsa Dadar1,2,10, M Mallar Chakravarty1,2,10,14
ORCID iDs: Olivier Parent, Aurélie Bussy, Manuela Costantino, Jérémie P Fouquet, Daniela Quesada, Alexa Pichet-Binette, Gabriel A Devenyi, Christopher J Steele, Sylvia Villeneuve
14 affiliations
  1. Cerebral Imaging Centre, Douglas Mental Health University Institute, Montreal, Canada
  2. Integrated Program in Neuroscience, McGill University, Montreal, Canada
  3. Center for the Studies in the Prevention of Alzheimer’s Disease, Douglas Mental Health University Institute, Montreal, Canada
  4. Centre d’études avancées en médecine du sommeil, Université de Montréal, Montreal, Canada
  5. Département de pharmacologie et physiologie, Université de Montréal, Montreal, Canada
  6. Centre de recherche de l’Institut universitaire de gériatrie de Montréal, Montreal, Canada
  7. Department of Anatomy and Cell Biology, McGill University, Montreal, Canada
  8. Département d’Anatomie, Université du Québec à Trois-Rivières, Trois-Rivières, Canada
  9. Department of Neurology and Neurosurgery, McGill University, Montreal, Canada
  10. Department of Psychiatry, McGill University, Montreal, Canada
  11. Department of Psychology, Concordia University, Montreal, Canada
  12. Department of Neurology, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany
  13. McConnell Brain Imaging Centre, Montreal Neurological Institute, Montreal, Canada
  14. Department of Biomedical Engineering, McGill University, Montreal, Canada
Institutions: Douglas Mental Health University Institute (Canada); McGill University (Canada); Canadian Sleep & Circadian Network (Canada); Université de Montréal (Canada); Institut Universitaire de Gériatrie de Montréal (Canada); Université du Québec à Trois-Rivières (Canada); Max Planck Institute for Human Cognitive and Brain Sciences (Germany); Concordia University (Canada); Montreal Neurological Institute and Hospital (Canada)
Journal: Nature communications, volume 17, issue 1, article 4623
Dates: received 22 July 2025; accepted 5 March 2026; published online 31 March 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1038/s41467-026-70832-2 · PMID 41916976 · PMCID PMC13199390 · OpenAlex W7143945179
Open access: gold, a free copy (OpenAlex)
Status: code verified (how the code is found and verified)
Categories: structural MRI / diffusion (modality), human (organism), other condition (population), Alzheimer's / dementia (population), stroke (population)
Methods: Connectivity, Statistics, Smoothing, state filtering, decompositions, Machine learning, Preprocessing, fMRI & imaging
Keywords: Alzheimer's disease, White matter disease
MeSH: Alzheimer Disease*, White Matter*, Aged, Aged, 80 and over, Brain, Dementia, Female, Humans, Magnetic Resonance Imaging, Male, Neurodegenerative Diseases (* major topic)
Topic: Dementia and Cognitive Impairment Research (Psychiatry and Mental health, Medicine), according to OpenAlex
Funding: Alzheimer Society; Fonds de Recherche du Québec - Santé
Citations: cited by 4 papers (Europe PMC); 96 references in the paper

Abstract

White matter hyperintensities (WMHs) are neuroimaging markers widely interpreted as caused by cerebral small vessel disease, yet emerging evidence suggests that a subset may have a neurodegenerative etiology. Current imaging methods have lacked the specificity to disentangle biological processes underlying WMHs in vivo. Here, we used voxel-level normative modeling and seven microstructural MRI markers with complementary biophysical sensitivities to generate single-subject high-resolution WMH pathophysiology maps in a large cohort (n = 32,526). We calculated data-driven spatial patterns of similar WMHs, revealing distinct periventricular, posterior, and anterior clusters. We identified a reproducible WMH signature linked to dementia and Alzheimer’s disease, characterized by a posterior predominance and a pathophysiological pattern indicative of selective fiber degeneration. Posterior WMHs connected cortical regions vulnerable to tau pathology. Our framework helps parsing vascular and neurodegenerative contributions of WMHs in vivo, which could alter the course of treatment strategies and provide nuanced interpretations of research findings.

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 25 matches between paragraphs and lines of code.

CoBrALab/WMH_patho_UKB

License: none: the authors keep all their rights
State: the link answers, verified on 30 September 2026
Evidence: files inventoried
Commit: ad8321861538cb467f4388729847222963f9067f, 11 February 2026
Languages: R (31), Python (5)
Size: 383 files, 36 scripts
Software Heritage: archived
Found in: “Code availability”
Holds: README
Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: data.table (31 files), tidyverse (28 files), ggplot2 (27 files), patchwork (16 files), cowplot (8 files), Matplotlib (5 files), NumPy (5 files), pandas (5 files), scikit-learn (4 files), broom (2 files), ggpubr (2 files), reshape2 (2 files), seaborn (2 files), emmeans (1 file), lme4 (1 file), lmerTest (1 file), multcomp (1 file), neuroHarmonize (1 file), neuromaps (1 file), NiBabel (1 file), Nilearn (1 file), Plotly (1 file), rstatix (1 file), SciPy (1 file), statsmodels (1 file)
Availability: 1 check, the latest on 30 September 2026: the link answers
  • 30 September 2026: the link answers
37 files
At the source: github.com/CoBrALab/WMH_patho_UKB

cobralab/documentation

License: none: the authors keep all their rights
State: the link answers, verified on 31 October 2026
Evidence: files inventoried
Commit: 1c7422abf8ac42c2796381a4c2c0b568f1bbf8c1, 15 January 2026
Size: 22 files, 0 scripts
Software Heritage: not archived
Found in: the text, “Quality control”
Holds: README, environment (pyproject.toml, uv.lock), continuous integration
Not found: license file, CITATION.cff, tests, documentation
Availability: 2 checks, the latest on 31 October 2026: the link answers
  • 31 October 2026: the link answers
  • 29 September 2026: the link answers
1 file
At the source: github.com/CoBrALab/documentation/wiki/Motion-Quality-Control-(QC)-Manual

Zenodo 18613580

License: CC-BY-4.0
State: the link answers, verified on 30 September 2026
Evidence: files inventoried
Size: 1 file
Software Heritage: not checked
Found in: the references
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Availability: 1 check, the latest on 30 September 2026: the link answers (HTTP 200)
  • 30 September 2026: the link answers (HTTP 200)
At the source:

Code availability

We share all analysis code, results, and figures in raw form, and our derived UKB template space and WMH parcellation based on pathophysiology on GitHub (https://github.com/CoBrALab/WMH_patho_UKB)94.

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

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:

  • 3 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 36 scripts, each with its path and the digest of its content;
  • 25 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 availability

Data from the UK Biobank is publicly available via material transfer agreements (https://www.ukbiobank.ac.uk/enable-your-research/register). Data from ADNI is publicly available via a data-sharing application (https://adni.loni.usc.edu/data-samples/adni-data/#AccessData). Data from PREVENT-AD is publicly available via a data user agreement (https://openpreventad.loris.ca/). Source data are provided in this paper.

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, 30 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 19 authors, 2 keywords, 11 MeSH terms, 2 funders, 93 references.

Cite

This paper

Parent, O., Alasmar, Z., Osborne, S., Bussy, A., Costantino, M., Fouquet, J. P., Quesada, D., Pastor-Bernier, A., Fajardo-Valdez, A., Pichet-Binette, A., McQuarrie, A., Maranzano, J., Devenyi, G. A., Steele, C. J., Villeneuve, S., the PREVENT-AD Research Group, the Alzheimer’s Disease Neuroimaging Initiative (ADNI), Dadar, M., & Chakravarty, M. M. (2026). Characterizing spatiotemporal white matter hyperintensity pathophysiology in vivo to disentangle vascular and neurodegenerative contributions. Nature communications, 17(1), 4623. https://doi.org/10.1038/s41467-026-70832-2

BibTeX

@article{parent2026characterizing,
author = {Parent, Olivier and Alasmar, Zaki and Osborne, Sophia and Bussy, Aurélie and Costantino, Manuela and Fouquet, Jérémie P and Quesada, Daniela and Pastor-Bernier, Alexandre and Fajardo-Valdez, Alfonso and Pichet-Binette, Alexa and McQuarrie, Ann and Maranzano, Josefina and Devenyi, Gabriel A and Steele, Christopher J and Villeneuve, Sylvia and {the PREVENT-AD Research Group} and {the Alzheimer’s Disease Neuroimaging Initiative (ADNI)} and Dadar, Mahsa and Chakravarty, M Mallar},
title = {{Characterizing spatiotemporal white matter hyperintensity pathophysiology in vivo to disentangle vascular and neurodegenerative contributions}},
journal = {Nature communications},
year = {2026},
month = mar,
volume = {17},
number = {1},
pages = {4623},
publisher = {Nature Publishing Group},
issn = {2041-1723},
doi = {10.1038/s41467-026-70832-2},
url = {https://doi.org/10.1038/s41467-026-70832-2},
pmid = {41916976},
pmcid = {PMC13199390}
}

RIS

TY - JOUR
AU - Parent, Olivier
AU - Alasmar, Zaki
AU - Osborne, Sophia
AU - Bussy, Aurélie
AU - Costantino, Manuela
AU - Fouquet, Jérémie P
AU - Quesada, Daniela
AU - Pastor-Bernier, Alexandre
AU - Fajardo-Valdez, Alfonso
AU - Pichet-Binette, Alexa
AU - McQuarrie, Ann
AU - Maranzano, Josefina
AU - Devenyi, Gabriel A
AU - Steele, Christopher J
AU - Villeneuve, Sylvia
AU - the PREVENT-AD Research Group
AU - the Alzheimer’s Disease Neuroimaging Initiative (ADNI)
AU - Dadar, Mahsa
AU - Chakravarty, M Mallar
TI - Characterizing spatiotemporal white matter hyperintensity pathophysiology in vivo to disentangle vascular and neurodegenerative contributions
T2 - Nature communications
J2 - Nat Commun
PY - 2026
DA - 2026/03/31
VL - 17
IS - 1
SP - 4623
SN - 2041-1723
PB - Nature Publishing Group
DO - 10.1038/s41467-026-70832-2
UR - https://doi.org/10.1038/s41467-026-70832-2
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s41467-026-70832-2",
"type": "article-journal",
"title": "Characterizing spatiotemporal white matter hyperintensity pathophysiology in vivo to disentangle vascular and neurodegenerative contributions",
"container-title": "Nature communications",
"author": [
{
"family": "Parent",
"given": "Olivier"
},
{
"family": "Alasmar",
"given": "Zaki"
},
{
"family": "Osborne",
"given": "Sophia"
},
{
"family": "Bussy",
"given": "Aurélie"
},
{
"family": "Costantino",
"given": "Manuela"
},
{
"family": "Fouquet",
"given": "Jérémie P"
},
{
"family": "Quesada",
"given": "Daniela"
},
{
"family": "Pastor-Bernier",
"given": "Alexandre"
},
{
"family": "Fajardo-Valdez",
"given": "Alfonso"
},
{
"family": "Pichet-Binette",
"given": "Alexa"
},
{
"family": "McQuarrie",
"given": "Ann"
},
{
"family": "Maranzano",
"given": "Josefina"
},
{
"family": "Devenyi",
"given": "Gabriel A"
},
{
"family": "Steele",
"given": "Christopher J"
},
{
"family": "Villeneuve",
"given": "Sylvia"
},
{
"literal": "the PREVENT-AD Research Group"
},
{
"literal": "the Alzheimer’s Disease Neuroimaging Initiative (ADNI)"
},
{
"family": "Dadar",
"given": "Mahsa"
},
{
"family": "Chakravarty",
"given": "M Mallar"
}
],
"container-title-short": "Nat Commun",
"volume": "17",
"issue": "1",
"page": "4623",
"DOI": "10.1038/s41467-026-70832-2",
"PMID": "41916976",
"PMCID": "PMC13199390",
"ISSN": "2041-1723",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s41467-026-70832-2",
"language": "en",
"issued": {
"date-parts": [
[
2026,
3,
31
]
]
}
}

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:10.1093/braincomms/fcad279 [code]
Assessment of white matter hyperintensity severity using multimodal magnetic resonance imaging.
Journal: Brain communications
In common: multcomp, broom, ggplot2, 7 other tools, stroke, Alzheimer's / dementia, structural MRI / diffusion, 18 references, 3 authors
[2] doi:10.1093/braincomms/fcaf475 [code]
Investigating links between white matter hyperintensities and menopausal status using robust age-correction methods in UK Biobank.
Journal: Brain communications
In common: lmerTest, lme4, data.table, 2 other tools, ukbiobank.ac.uk/enable-your-research/register, stroke, structural MRI / diffusion, 9 references, 2 authors
[3] doi:10.1038/s41593-024-01763-8 [code]
Synergistic association of Aβ and tau pathology with cortical neurophysiology and cognitive decline in asymptomatic older adults.
Journal: Nature neuroscience
In common: multcomp, broom, rstatix, 7 other tools, Alzheimer's / dementia, 3 references, 2 authors
[4] doi:10.1038/s41598-025-21596-0 [code]
Longitudinal functional connectivity during rest and task is differentially related to Alzheimer's pathology and episodic memory in older adults.
Journal: Scientific reports
In common: rstatix, lmerTest, Plotly, 8 other tools, Alzheimer's / dementia, 2 references, 2 authors
[5] doi:10.1523/jneurosci.1408-24.2024 [code]
Precuneus Activity during Retrieval Is Positively Associated with Amyloid Burden in Cognitively Normal Older <i>APOE</i>4 Carriers.
Journal: The Journal of neuroscience : the official journal of the Society for Neuroscience
In common: emmeans, lmerTest, Plotly, 8 other tools, Alzheimer's / dementia, 2 references, 2 authors
[6] doi:10.1093/braincomms/fcaf432 [code]
The effect of chronic stress and chronic alcohol intake on behaviour, brain volume, and functional connectivity in a longitudinal rat model.
Journal: Brain communications
In common: multcomp, broom, rstatix, 11 other tools, structural MRI / diffusion, other condition, author Gabriel A. Devenyi
[7] doi:10.1212/wnl.0000000000218472 [code]
Lesion-Level Subtypes of White Matter Hyperintensity Evolution Beyond Spatial Location.
Journal: Neurology
In common: Nilearn, Plotly, statsmodels, 6 other tools, stroke, Alzheimer's / dementia, structural MRI / diffusion, 10 references
[8] doi:10.1002/hbm.26259 [code]
High spatial overlap but diverging age-related trajectories of cortical magnetic resonance imaging markers aiming to represent intracortical myelin and microstructure.
Journal: Human brain mapping
In common: reshape2, ggplot2, structural MRI / diffusion, 5 references, 3 authors
[9] doi:10.1002/alz.70196 [code]
Cognitive implications and associated transcriptomic signatures of distinct regional iron depositions in cerebral small vessel disease.
Journal: Alzheimer's & dementia : the journal of the Alzheimer's Association
In common: neuromaps, Nilearn, Plotly, 13 other tools, stroke, Alzheimer's / dementia, structural MRI / diffusion, 3 references
[10] doi:10.1038/s42003-026-10066-6 [code]
Morphological and anatomical variations in subcortical anatomy between humans and chimpanzees associated with heritability patterns related to human behavioral traits.
Journal: Communications biology
In common: seaborn, scikit-learn, pandas, 2 other tools, 2 references, 3 authors

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.

Add the badge to the code's README

OSCR: the badge of this paper

If you are an author or a maintainer of this paper's code, you can add this badge. It says what OSCR knows of the paper (its code, its validated tracing map) and links the code to this page. Paste one of these snippets under the README's title. Keep the two marks around it, so that the badge can be replaced in place and is never added twice. OSCR changes nothing in your repository.

Markdown
<!-- oscr-badge:start -->
[![OSCR](https://openscicode.org/badge/doi/10.1038/s41467-026-70832-2.svg)](https://openscicode.org/badge/doi/10.1038/s41467-026-70832-2)
<!-- oscr-badge:end -->
reStructuredText
.. oscr-badge:start

.. image:: https://openscicode.org/badge/doi/10.1038/s41467-026-70832-2.svg
   :target: https://openscicode.org/badge/doi/10.1038/s41467-026-70832-2
   :alt: OSCR

.. oscr-badge:end
HTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.1038/s41467-026-70832-2"><img src="https://openscicode.org/badge/doi/10.1038/s41467-026-70832-2.svg" alt="OSCR"></a>
<!-- oscr-badge:end -->

On GitHub, open the README, choose Edit (the pencil), paste the snippet under the title, then "Commit changes".

If you choose to, you can have the badge kept up to date for you, with oscr badge --write from the command line or with the GitHub Action in your repository's workflows. Every other form and option (a card, AsciiDoc, LaTeX, shields.io) is on the badge 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: code errors and mismatches between the code and the paper about this paper and its code, from readers and its authors, beside the paper and its tracing map.

Report a mismatch between the code and the paper · Report a code error (signed in)

Reproductions: reports from readers who ran the authors' code and did not get the paper's result, what they ran, with which environment, commit and data.

Report a reproduction failure (signed in)

Versions of its code: the releases of its code tied to a version of this paper (the preprint, the accepted manuscript, the version of record…), each with the tracing map versioned with it.

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