OSCR

Modelling variability in functional brain networks using embeddings.

Code ↔ Paper

12 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 12 matches
  1. [1] § Methods › Datasets › Real MEG data ↔ data_preprocessing/wakeman_henson/1_preprocess.py, lines 11–44 · score 0.86 · notch filter, bad segment, Wakeman Henson, 125 Hz, ECG, EoG
  2. [2] § Methods › Datasets › Real MEG data ↔ data_preprocessing/notts/1_preprocess.py, lines 14–52 · score 0.85 · notch filter, bad segment, bad channel, 125 Hz, ECG, EoG
  3. [3] § Results › Real MEG data › Wakeman Henson: HIVE reveals similarities and differences between MEG recordings ↔ real_data/wakeman_henson/analyse_and_plot.py, lines 253–311 · score 0.78 · Calinski Harabasz score, Davies Bouldin score, Silhouette score, HMM DE, clusters, dual
  4. [4] § Results › Real MEG data › Wakeman Henson: HIVE reveals similarities and differences between MEG recordings ↔ real_data/multi_dataset/analyse_and_plot.py, lines 483–551 · score 0.77 · Calinski Harabasz score, Davies Bouldin score, Silhouette score, HMM DE, clusters, dual
  5. [5] § Results › Real MEG data › Wakeman Henson: HIVE reveals similarities and differences between MEG recordings ↔ real_data/wakeman_henson/analyse_and_plot.py, lines 253–311 · score 0.74 · Calinski Harabasz score, negative Davies Bouldin, Silhouette score, Wakeman Henson, HMM DE, Clustering
  6. [6] § Results › Real MEG data › Wakeman Henson: HIVE reveals similarities and differences between MEG recordings ↔ real_data/multi_dataset/analyse_and_plot.py, lines 483–551 · score 0.72 · Calinski Harabasz score, negative Davies Bouldin, Silhouette score, HMM DE, Clustering, HIVE
  7. [7] § Results › Real MEG data › Cam-CAN: HIVE reveals individual variability across age ↔ real_data/camcan/predict_age.py, lines 87–147 · score 0.62 · cross validation, predict age, folds, demographics, dimensions, training
  8. [8] § Methods › Datasets › Real MEG data ↔ data_preprocessing/notts/3_source_reconstruct.py, lines 88–129 · score 0.61 · Source reconstruction, surfaces, beamformer, nose, symmetric, Coregistration
  9. [9] § Methods › Datasets › Real MEG data ↔ data_preprocessing/wakeman_henson/2_source_reconstruct.py, lines 34–81 · score 0.60 · Source reconstruction, surfaces, beamformer, nose, symmetric, Coregistration
  10. [10] § Results › Real MEG data › Cam-CAN: HIVE reveals individual variability across age ↔ real_data/camcan/hive.py, lines 92–226 · score 0.58 · nearest neighbours, power maps, young, Centroids, PSDs, age
  11. [11] § Results › Real MEG data › Cam-CAN: HIVE reveals individual variability across age ↔ real_data/camcan/analyse_and_plot.py, lines 258–368 · score 0.57 · nearest neighbours, power maps, young, Centroids, PSDs, age
  12. [12] § Results › Real MEG data › Cam-CAN: HIVE reveals individual variability across age ↔ real_data/camcan/predict_age.py, lines 87–147 · score 0.55 · age prediction, cross validated, fold, PCA, dimensions, HMM

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

Python · 347 lines · 9.7 KB · MIT · 2 matches

  1. import os
  2. from glob import glob
  3. from collections import defaultdict
  5. import numpy as np
  6. import pickle
  7. import matplotlib.pyplot as plt
  8. from scipy.stats import ttest_ind
  9. from scipy.spatial.distance import pdist, squareform
  10. from sklearn.metrics import (
  11. davies_bouldin_score,
  12. silhouette_score,
  13. calinski_harabasz_score,
  14. )
  16. from osl_dynamics.utils import plotting, set_random_seed
  17. from osl_dynamics.inference import tf_ops
  18. from osl_dynamics.models import hive
  19. from osl_dynamics.data import Data
  20. from osl_dynamics.analysis import spectral, power, connectivity
  22. tf_ops.gpu_growth()
  24. set_random_seed(0)
  27. def get_best_dimension(dimensions, significance_level=0.01):
  28. best_dim = dimensions[0]
  29. fe_1 = [
  30. pickle.load(open(f"results/hive_{best_dim}/run{i}/model/history.pkl", "rb"))[
  31. "free_energy"
  32. ]
  33. for i in range(1, 11)
  34. ]
  36. for i in range(len(dimensions) - 1):
  37. dim = dimensions[i + 1]
  38. fe_2 = [
  39. pickle.load(open(f"results/hive_{dim}/run{i}/model/history.pkl", "rb"))[
  40. "free_energy"
  41. ]
  42. for i in range(1, 11)
  43. ]
  44. pvalue = ttest_ind(fe_1, fe_2, alternative="greater", permutations=5000).pvalue
  45. print(f"p-value for {best_dim} vs {dim}: {pvalue}")
  46. if pvalue < significance_level:
  47. best_dim = dim
  48. fe_1 = fe_2
  50. print(f"Best dimension: {best_dim}")
  51. return best_dim
  54. def get_best_hive_run(dim):
  55. best_fe = np.Inf
  56. for run in range(1, 11):
  57. history = pickle.load(
  58. open(f"results/hive_{dim}/run{run}/model/history.pkl", "rb")
  59. )
  60. if history["loss"][-1] < best_fe:
  61. best_fe = history["loss"][-1]
  62. best_run = run
  64. print(f"Best HIVE run: {best_run}")
  65. return best_run
  68. def get_best_hmm_run():
  69. best_fe = np.Inf
  70. for run in range(1, 11):
  71. history = pickle.load(open(f"results/hmm/run{run}/model/history.pkl", "rb"))
  72. if history["loss"][-1] < best_fe:
  73. best_fe = history["loss"][-1]
  74. best_run = run
  76. print(f"Best HMM run: {best_run}")
  77. return best_run
  80. def load_data(use_tfrecord=True, buffer_size=2000, n_jobs=16):
  81. """Load the data."""
  83. data_paths = sorted(
  84. glob(
  85. "/well/woolrich/projects/wakeman_henson/spring23/src/sub*_run*/sflip_parc-raw.fif"
  86. )
  87. )
  88. training_data = Data(
  89. data_paths,
  90. sampling_frequency=250,
  91. mask_file="MNI152_T1_8mm_brain.nii.gz",
  92. parcellation_file="fmri_d100_parcellation_with_PCC_reduced_2mm_ss5mm_ds8mm.nii.gz",
  93. picks="misc",
  94. reject_by_annotation="omit",
  95. use_tfrecord=use_tfrecord,
  96. buffer_size=buffer_size,
  97. n_jobs=n_jobs,
  98. )
  100. return training_data
  103. best_dim = get_best_dimension([3, 5, 10, 20, 30])
  104. best_hive_run = get_best_hive_run(best_dim)
  105. hive_loss = pickle.load(
  106. open(f"results/hive_{best_dim}/run{best_hive_run}/model/history.pkl", "rb")
  107. )["loss"]
  109. best_hmm_run = get_best_hmm_run()
  110. hmm_loss = pickle.load(open(f"results/hmm/run{best_hmm_run}/model/history.pkl", "rb"))[
  111. "loss"
  112. ]
  113. model = hive.Model.load(f"results/hive_{best_dim}/run{best_hive_run}/model")
  114. inf_params_dir = f"results/hive_{best_dim}/run{best_hive_run}/inf_params"
  116. best_run_dir = f"results/best_run"
  117. os.makedirs(best_run_dir, exist_ok=True)
  119. plot_dir = f"{best_run_dir}/plots"
  120. os.makedirs(plot_dir, exist_ok=True)
  122. plotting.plot_line(
  123. [range(len(hive_loss)), range(len(hmm_loss))],
  124. [hive_loss, hmm_loss],
  125. labels=["HIVE", "HMM-DE"],
  126. filename=f"{plot_dir}/loss.png",
  127. )
  129. alpha = pickle.load(open(f"{inf_params_dir}/alp.pkl", "rb"))
  130. embeddings = np.load(f"{inf_params_dir}/summed_embeddings.npy")
  132. plotting.plot_alpha(
  133. alpha[0],
  134. n_samples=2000,
  135. cmap="tab10",
  136. filename=f"{plot_dir}/alpha.png",
  137. )
  139. # Pairwise correlation between subject embeddings
  140. se_cosine = squareform(pdist(embeddings, metric="cosine"))
  141. fig, ax = plotting.plot_matrices(
  142. se_cosine,
  143. cmap="coolwarm",
  144. )
  145. ax[0][0].set_xticks(
  146. ticks=np.arange(0, 114, 6) + 3, labels=[f"{i + 1}" for i in range(19)]
  147. )
  148. ax[0][0].set_yticks(
  149. ticks=np.arange(0, 114, 6) + 3, labels=[f"{i + 1}" for i in range(19)]
  150. )
  151. fig.savefig(f"{plot_dir}/embeddings_cosine.png")
  154. training_data = load_data(n_jobs=16)
  155. trimmed_data = training_data.trim_time_series(
  156. sequence_length=200,
  157. n_embeddings=15,
  158. prepared=False,
  159. )
  160. spectra = spectral.multitaper_spectra(
  161. data=trimmed_data,
  162. alpha=alpha,
  163. sampling_frequency=250,
  164. n_jobs=16,
  165. )
  166. pickle.dump(spectra, open(f"{inf_params_dir}/spectra.pkl", "wb"))
  168. f, psd, coh = pickle.load(open(f"{inf_params_dir}/spectra.pkl", "rb"))
  169. nnmf = spectral.decompose_spectra(coh, n_components=2)
  170. np.save(f"{inf_params_dir}/nnmf_2.npy", nnmf)
  172. f, psd, coh = pickle.load(open(f"{inf_params_dir}/spectra.pkl", "rb"))
  173. frequency_range = [1, 45]
  174. n_components = nnmf.shape[0]
  175. plotting.plot_line(
  176. [f] * n_components,
  177. nnmf,
  178. labels=[f"Component {i}" for i in range(n_components)],
  179. x_label="Frequency (Hz)",
  180. y_label="Weighting",
  181. )
  183. # Calculate group average
  184. gpsd = np.average(psd, axis=0)
  185. gcoh = np.average(coh, axis=0)
  187. # Calculate average PSD across channels and the standard error
  188. p = np.mean(gpsd, axis=-2)
  189. e = np.std(gpsd, axis=-2) / np.sqrt(gpsd.shape[-2])
  191. # Plot PSDs
  192. n_states = gpsd.shape[0]
  193. fig, axes = plt.subplots(1, 6, figsize=(36, 6))
  194. for i in range(model.config.n_states):
  195. plotting.plot_line(
  196. [f],
  197. [p[i]],
  198. errors=[[p[i] - e[i]], [p[i] + e[i]]],
  199. labels=[f"State {i + 1}"],
  200. x_range=[f[0], f[-1]],
  201. y_range=[p.min() - 0.1 * p.max(), 1.2 * p.max()],
  202. x_label="Frequency (Hz)",
  203. y_label="PSD (a.u.)",
  204. ax=axes[i],
  205. )
  206. axes[i].axvspan(
  207. frequency_range[0],
  208. frequency_range[1],
  209. alpha=0.25,
  210. color="gray",
  211. )
  212. fig.savefig(f"{plot_dir}/spectra_psd.png")
  214. gp = power.variance_from_spectra(f, gpsd, nnmf)
  215. power.save(
  216. gp,
  217. mask_file=training_data.mask_file,
  218. parcellation_file=training_data.parcellation_file,
  219. subtract_mean=True,
  220. show_plots=False,
  221. filename=f"{plot_dir}/psd.png",
  222. combined=True,
  223. titles=[f"Mode {i+1}" for i in range(model.config.n_states)],
  224. plot_kwargs={"views": ["lateral"]},
  225. )
  227. gc = connectivity.mean_coherence_from_spectra(f, gcoh, nnmf)
  228. gc = connectivity.threshold(gc, percentile=97, subtract_mean=True)
  229. connectivity.save(
  230. gc,
  231. parcellation_file=training_data.parcellation_file,
  232. combined=True,
  233. titles=[f"Mode {i+1}" for i in range(model.config.n_states)],
  234. filename=f"{plot_dir}/coh.png",
  235. )
  237. clustering_scores = {
  238. "hmm": defaultdict(list),
  239. "hive": defaultdict(list),
  240. }
  241. subject_labels = np.repeat(np.arange(19), 6)
  243. for model in ["hmm", "hive"]:
  244. model_dir = f"results/{model}"
  245. for run in range(1, 11):
  246. if model == "hive":
  247. covs = np.load(
  248. f"{model_dir}_{best_dim}/run{run}/inf_params/session_covs.npy"
  249. )
  250. else:
  251. covs = np.load(f"{model_dir}/run{run}/dual_estimates/covs.npy")
  253. covs_flatten = np.array([cov.flatten() for cov in covs])
  254. clustering_scores[model]["silhouette"].append(
  255. silhouette_score(covs_flatten, subject_labels)
  256. )
  257. clustering_scores[model]["davies_bouldin"].append(
  258. davies_bouldin_score(covs_flatten, subject_labels)
  259. )
  260. clustering_scores[model]["calinski_harabasz"].append(
  261. calinski_harabasz_score(covs_flatten, subject_labels)
  262. )
  264. plotting.plot_violin(
  265. np.array(
  266. [
  267. clustering_scores["hive"]["silhouette"],
  268. clustering_scores["hmm"]["silhouette"],
  269. ]
  270. ),
  271. ["HIVE", "HMM-DE"],
  272. title="Silhouette score",
  273. y_label="Score",
  274. sns_kwargs={"cut": 0, "scale": "width"},
  275. filename=f"{plot_dir}/silhouette_scores.png",
  276. )
  277. plotting.plot_violin(
  278. 1
  279. - np.array(
  280. [
  281. clustering_scores["hive"]["davies_bouldin"],
  282. clustering_scores["hmm"]["davies_bouldin"],
  283. ]
  284. ),
  285. ["HIVE", "HMM-DE"],
  286. title="Negative Davies-Bouldin score",
  287. y_label="Score",
  288. sns_kwargs={"cut": 0, "scale": "width"},
  289. filename=f"{plot_dir}/davies_bouldin_scores.png",
  290. )
  291. plotting.plot_violin(
  292. np.array(
  293. [
  294. clustering_scores["hive"]["calinski_harabasz"],
  295. clustering_scores["hmm"]["calinski_harabasz"],
  296. ]
  297. ),
  298. ["HIVE", "HMM-DE"],
  299. title="Calinski-Harabasz score",
  300. y_label="Score",
  301. sns_kwargs={"cut": 0, "scale": "width"},
  302. filename=f"{plot_dir}/calinski_harabasz_scores.png",
  303. )
  305. best_hmm_covs = np.load(f"results//hmm/run{best_hmm_run}/dual_estimates/covs.npy")
  306. best_hive_covs = np.load(
  307. f"results/hive_{best_dim}/run{best_hive_run}/inf_params/session_covs.npy"
  308. )
  310. best_hmm_covs_flatten = np.array([cov.flatten() for cov in best_hmm_covs])
  311. best_hive_covs_flatten = np.array([cov.flatten() for cov in best_hive_covs])
  313. # Get the pairwise distances
  314. hmm_pdist = squareform(pdist(best_hmm_covs_flatten, metric="euclidean"))
  315. hive_pdist = squareform(pdist(best_hive_covs_flatten, metric="euclidean"))
  317. fig, ax = plotting.plot_matrices(
  318. [
  319. hive_pdist,
  320. hmm_pdist,
  321. ],
  322. titles=["HIVE", "HMM-DE"],
  323. cmap="coolwarm",
  324. )
  325. ax[0][0].set_xticks(
  326. ticks=np.arange(0, 114, 6) + 3,
  327. labels=[f"{i + 1}" for i in range(19)],
  328. fontsize=6,
  329. )
  330. ax[0][0].set_yticks(
  331. ticks=np.arange(0, 114, 6) + 3,
  332. labels=[f"{i + 1}" for i in range(19)],
  333. fontsize=6,
  334. )
  335. plt.setp(ax[0][0].get_xticklabels(), rotation=45)
  336. ax[0][1].set_xticks(
  337. ticks=np.arange(0, 114, 6) + 3,
  338. labels=[f"{i + 1}" for i in range(19)],
  339. fontsize=6,
  340. )
  341. ax[0][1].set_yticks(
  342. ticks=np.arange(0, 114, 6) + 3,
  343. labels=[f"{i + 1}" for i in range(19)],
  344. fontsize=6,
  345. )
  346. plt.setp(ax[0][1].get_xticklabels(), rotation=45)
  347. fig.savefig(f"{plot_dir}/covs_pairwise_distances.png")

analyse_and_plot.py at commit 7456b9a, under MIT · at the source

Overview

Authors: Rukuang Huang1, Chetan Gohil1, Mark Woolrich1
ORCID iDs: Rukuang Huang
  1. Oxford Centre for Human Brain Activity (OHBA), Wellcome Centre for Integrative Neuroimaging, Department of Psychiatry, University of Oxford, Oxford, United Kingdom
Institutions: University of Oxford (United Kingdom); Wellcome Centre for Integrative Neuroimaging (United Kingdom)
Journal: Imaging neuroscience (Cambridge, Mass.), volume 4, article IMAG.a.1188
Dates: received 31 July 2024; accepted 24 February 2026; published online 17 April 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1162/imag.a.1188 · PMID 42016560 · PMCID PMC13094015 · OpenAlex W7135035904
Open access: diamond, a free copy (OpenAlex)
Status: code verified
Categories: systems (subfield)
Methods: Spectral & time-frequency, Preprocessing, Connectivity, Statistics, Smoothing, state filtering, decompositions, Machine learning
Keywords: functional connectivity, population modelling, unsupervised learning, generative modelling, Bayesian modelling
Topic: Functional Brain Connectivity Studies (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: Engineering and Physical Sciences Research Council (EP/S02428X/1); Wellcome Trust (215573/Z/19/Z, 106183/Z/14/Z); NIHR Oxford Biomedical Research Centre (NIHR203316)
Citations: cited by 4 papers (Europe PMC); 69 references in the paper

Abstract

Functional neuroimaging techniques allow us to estimate functional networks that underlie cognition. However, these functional networks are often estimated at the group level and do not allow for the discovery of, nor benefit from, subpopulation structure in the data, that is, the fact that some recording sessions may be more similar than others. Here, we propose the use of embedding vectors (c.f. word embedding in Natural Language Processing) to explicitly model individual sessions while inferring networks across a group. This vector is effectively a “fingerprint” for each session, which can cluster sessions with similar functional networks together in a learnt embedding space. We apply this approach to estimate dynamic functional networks using a hierarchical Hidden Markov Model (HMM). We call this approach HIVE (HMM with Integrated Variability Estimation). Using simulated data, we show that HIVE can uncover true subpopulation structure and show improved performance over existing approaches. Using real magnetoencephalography data, we show the learnt embedding vectors (session fingerprints) reflect meaningful sources of variation across a population. Overall, HIVE provides a powrful new approach for modelling individual sessions while leveraging information available across an entire group.

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

Repository

Its files are read in the Code ↔ Paper reader above, with 12 matches between paragraphs and lines of code.

OHBA-analysis/Huang2025_ModelVariabilityWithEmbeddings

License: MIT
State: the link answers, verified on 29 September 2026
Evidence: files inventoried
Commit: 7456b9af6601e253f6728823e2c99b14e2d3e9ec, 26 February 2025
Languages: Python (27)
Size: 32 files, 27 scripts
Software Heritage: not archived
Found in: the text, “Results”
Holds: README, license file
Not found: CITATION.cff, environment file, tests, continuous integration, documentation
Tools: NumPy (14 files), OHBA Software Library (OSL) (11 files), Matplotlib (8 files), pandas (8 files), scikit-learn (7 files), SciPy (6 files), seaborn (6 files), NiBabel (1 file)
Availability: 1 check, the latest on 29 September 2026: the link answers
  • 29 September 2026: the link answers
29 files
At the source: github.com/OHBA-analysis/Huang2025_ModelVariabilityWithEmbeddings

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:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 27 scripts, each with its path and the digest of its content;
  • 12 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

No dataset and no data link were found in the paper.

Data and Code Availability

Data used are publicly available. Availability of the Nottingham dataset is at the official MEGUK site: https://meguk.ac.uk/database/. For the Wakeman-Henson dataset, we refer the readers to the original paper (Wakeman & Henson, 2015). For the Cam-CAN dataset, we refer the readers to the original paper (J. R. Taylor et al., 2017). Source code for HIVE is available in the osl-dynamics toolbox (Gohil et al., 2023) and scripts to reproduce results in this paper are available here:

github.com/OHBA-analysis/Huang2025_ModelVariabilityWithEmbeddings.

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, 29 September 2026: the first record

Recorded: type, language, journal, volume, pages, dates, 3 authors, 5 keywords, 3 funders, 66 references.

Cite

This paper

Huang, R., Gohil, C., & Woolrich, M. (2026). Modelling variability in functional brain networks using embeddings. Imaging neuroscience (Cambridge, Mass.), 4, IMAG.a.1188. https://doi.org/10.1162/imag.a.1188

BibTeX

@article{huang2026modelling,
author = {Huang, Rukuang and Gohil, Chetan and Woolrich, Mark},
title = {{Modelling variability in functional brain networks using embeddings}},
journal = {Imaging neuroscience (Cambridge, Mass.)},
year = {2026},
month = apr,
volume = {4},
pages = {IMAG.a.1188},
publisher = {MIT Press},
issn = {2837-6056},
doi = {10.1162/imag.a.1188},
url = {https://doi.org/10.1162/imag.a.1188},
pmid = {42016560},
pmcid = {PMC13094015}
}

RIS

TY - JOUR
AU - Huang, Rukuang
AU - Gohil, Chetan
AU - Woolrich, Mark
TI - Modelling variability in functional brain networks using embeddings
T2 - Imaging neuroscience (Cambridge, Mass.)
J2 - Imaging Neurosci (Camb)
PY - 2026
DA - 2026/04/17
VL - 4
SP - IMAG.a.1188
SN - 2837-6056
PB - MIT Press
DO - 10.1162/imag.a.1188
UR - https://doi.org/10.1162/imag.a.1188
LA - en
ER -

CSL-JSON

{
"id": "10.1162/imag.a.1188",
"type": "article-journal",
"title": "Modelling variability in functional brain networks using embeddings",
"container-title": "Imaging neuroscience (Cambridge, Mass.)",
"author": [
{
"family": "Huang",
"given": "Rukuang"
},
{
"family": "Gohil",
"given": "Chetan"
},
{
"family": "Woolrich",
"given": "Mark"
}
],
"container-title-short": "Imaging Neurosci (Camb)",
"volume": "4",
"page": "IMAG.a.1188",
"DOI": "10.1162/imag.a.1188",
"PMID": "42016560",
"PMCID": "PMC13094015",
"ISSN": "2837-6056",
"publisher": "MIT Press",
"URL": "https://doi.org/10.1162/imag.a.1188",
"language": "en",
"issued": {
"date-parts": [
[
2026,
4,
17
]
]
}
}

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:10.1162/imag.a.1301 [code]
MEG-GPT: A transformer-based foundation model for magnetoencephalography data.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: NiBabel, seaborn, scikit-learn, 4 other tools, 15 references
[2] doi:10.1162/imag.a.1237 [code]
Modelling discrete states and long-term dynamics in functional brain networks.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: seaborn, scikit-learn, pandas, 3 other tools, systems, 12 references, author Rukuang Huang
[3] doi:10.1162/imag.a.1190 [code]
Canonical Hidden Markov Model Networks for studying M/EEG.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: OHBA Software Library (OSL), NiBabel, scikit-learn, 4 other tools, 12 references
[4] doi:10.1002/hbm.70516 [code]
Effects of Age on Resting-State Cortical Networks.
Journal: Human brain mapping
In common: OHBA Software Library (OSL), NiBabel, scikit-learn, 4 other tools, 10 references
[5] doi:10.1038/s41531-026-01372-1 [code]
Varying patterns of association between cortical large-scale networks and subthalamic nucleus activity in Parkinson's disease.
Journal: NPJ Parkinson's disease
In common: OHBA Software Library (OSL), NiBabel, seaborn, 5 other tools, systems, 6 references
[6] doi:10.1162/imag.a.1269 [code]
From early to contemporary normative modeling: Mapping individual differences in neurophysiological signals.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: NiBabel, seaborn, scikit-learn, 4 other tools, 4 references
[7] doi:10.1038/s41467-026-75745-8 [code]
A language network in the individualized functional connectomes of 1199 human brains doing arbitrary tasks.
Journal: Nature communications
In common: NiBabel, seaborn, scikit-learn, 4 other tools, 4 references
[8] doi:10.1093/braincomms/fcag236 [code]
Dynamic, state-dependent characteristics of cognitive fluctuations in Lewy body dementia: a magnetoencephalography study.
Journal: Brain communications
In common: OHBA Software Library (OSL), pandas, NumPy, 4 references
[9] doi:10.1038/s41467-026-76011-7 [code]
Human cortex organizes dynamic co-fluctuations along the sensorimotor-association axis.
Journal: Nature communications
In common: NiBabel, SciPy, Matplotlib, 1 other tool, systems, 4 references
[10] doi:10.7554/elife.108109 [code]
Multimodal MRI marker of cognition explains the association between cognition and mental health in the UK Biobank.
Journal: eLife
In common: seaborn, scikit-learn, pandas, 3 other tools, 3 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.