OSCR

Pretraining for Large-Scale Functional Connectome Fingerprinting Supports Generalization and Transfer Learning in Functional Neuroimaging.

Code ↔ Paper

5 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 5 matches
  1. [1] § Materials and Methods › Data and preprocessing ↔ preproc_model_embedding_code.zip/preproc_model_embedding_code/embedding_generation_script.py, lines 179–226 · score 0.81 · global signal, fMRIPrep, CSF, detrended, overlap, preprocessing
  2. [2] § Materials and Methods › Model architecture and pretraining ↔ preproc_model_embedding_code.zip/preproc_model_embedding_code/brain2vec_model_arch.py, lines 39–140 · score 0.75 · weight decay, class weighting, cross entropy, Adam, loss, epochs
  3. [3] § Materials and Methods › Data and preprocessing ↔ preproc_model_embedding_code.zip/preproc_model_embedding_code/embedding_generation_script.py, lines 44–48 · score 0.64 · minimal preprocessing, AOMIC PIOP1, fMRIPrep, derivatives
  4. [4] § Materials and Methods › Transfer-learning experiments: identification of new individuals ↔ preproc_model_embedding_code.zip/preproc_model_embedding_code/brain2vec_model_arch.py, lines 39–140 · score 0.62 · weight decay, cross entropy, momentum, model embedding, class, linear
  5. [5] § Materials and Methods › Model architecture and pretraining ↔ preproc_model_embedding_code.zip/preproc_model_embedding_code/brain2vec_model_arch.py, lines 149–258 · score 0.56 · PyTorch, ReLU, lightning, affine, leaky, momentum

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 · 358 lines · 13 KB · no license · 3 matches

  1. #!/usr/bin/env python3
  2. # -*- coding: utf-8 -*-
  4. import pytorch_lightning as pl
  5. from pytorch_lightning.callbacks.early_stopping import EarlyStopping
  6. import torchmetrics
  7. from sklearn.utils.class_weight import compute_class_weight
  8. import os
  9. from pathlib import Path
  10. import torch
  11. from torch.utils.data import Dataset, DataLoader
  12. import torch.nn as nn
  13. import torch.nn.functional as F
  14. import torch.optim as optim
  15. # import glob
  16. import numpy as np
  17. # import seaborn as sns
  18. import pandas as pd
  19. import matplotlib.pyplot as plt
  20. from tqdm import tqdm
  22. # this code might occasionally borrow code, approaches or generally draw inspiration from https://github.com/hltcoe/xvectors
  26. # Function to initialize model weights (called by the hltcoe xvec github code)
  27. def init_weight(model):
  29. for m in model.modules():
  30. if isinstance(m, nn.Linear) or isinstance(m, nn.Conv1d):
  31. nn.init.kaiming_normal_(m.weight, a=0.01) # default negative slope of LeakyReLU
  32. elif isinstance(m, nn.BatchNorm1d):
  33. if m.affine:
  34. nn.init.constant_(m.weight, 1)
  35. nn.init.constant_(m.bias, 0)
  39. class b2v_FFwdNN_uniform_layer_sizes(pl.LightningModule):
  40. def __init__(self, input_dim, n_layers: int = 4, layer_dim: int = 1024, embedding_dim: int = 1024, dropout_rate: float = 0.0, nclasses: int = 2,class_weights: list = [1,1], batch_size: int = 16, lr: float = 0.001):
  41. super().__init__()
  42. self.save_hyperparameters()
  44. self.batch_size = batch_size # can be tuned by pytorchlightning
  45. self.lr = lr # can be tuned by pytorchlightning
  47. self.n_layers = n_layers
  49. self.dropout_rate = dropout_rate
  51. self.input_dim = input_dim
  52. self.layer_dim = layer_dim
  54. self.embedding_dim = embedding_dim
  55. self.nclasses = nclasses
  56. self.class_weights = class_weights
  59. self.base_net = torch.nn.Sequential()
  60. self.base_net.add_module('layer1_linear', nn.Linear(self.input_dim, self.layer_dim))
  61. self.base_net.add_module('layer1_relu', nn.LeakyReLU(inplace=True))
  62. self.base_net.add_module('layer1_batchnorm', nn.BatchNorm1d(self.layer_dim, momentum=0.1,affine=False))
  63. self.base_net.add_module('layer1_dropout',nn.Dropout(p=self.dropout_rate))
  65. if self.n_layers > 1:
  66. for ix, ll in enumerate(list(range(1,self.n_layers))):
  67. self.base_net.add_module('layer'+ str(ll+1) + '_linear', nn.Linear(self.layer_dim, self.layer_dim))
  68. self.base_net.add_module('layer'+ str(ll+1) + '_relu', nn.LeakyReLU(inplace=True))
  69. self.base_net.add_module('layer'+ str(ll+1) + '_batchnorm', nn.BatchNorm1d(self.layer_dim, momentum=0.1,affine=False))
  70. self.base_net.add_module('layer'+ str(ll+1) + '_dropout',nn.Dropout(p=self.dropout_rate))
  72. self.embed1_linear = nn.Linear(self.layer_dim ,self.embedding_dim)
  73. self.embed1_relu = nn.LeakyReLU(inplace=True)
  74. self.embed1_batchnorm = nn.BatchNorm1d(self.embedding_dim, momentum=0.1,affine=False)
  76. self.output = nn.Linear(self.embedding_dim, self.nclasses)
  77. init_weight(self)
  79. self.softmax = nn.Softmax(dim=1);
  82. # Metrics
  83. self.train_accuracy = torchmetrics.Accuracy(num_classes=self.nclasses, task = "multiclass")
  85. self.val_accuracy = torchmetrics.Accuracy(num_classes=self.nclasses, task = "multiclass")
  88. def forward(self, x):
  90. base_out = self.base_net.forward(x.float())
  92. x = self.embed1_linear(base_out)
  93. x = self.embed1_relu(x)
  94. embed = self.embed1_batchnorm(x)
  96. out_logit = self.output(embed)
  98. out_softmax = self.softmax(out_logit)
  100. return out_logit, out_softmax, embed, base_out
  102. def configure_optimizers(self):
  103. optimizer = optim.Adam(self.parameters(), lr=self.lr, weight_decay=1e-4)
  104. lr_schedulers = {'scheduler': torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode='min',patience=3,verbose=True),
  105. 'monitor': 'val_loss'}
  106. return [optimizer], [lr_schedulers]
  108. def training_step(self, batch, batch_idx):
  109. x, y, y_str, f_path = batch
  110. output = self(x)
  111. y_hat_logits = output[0]
  112. y_hat_softmax = output[1]
  113. _, predicted = torch.max(y_hat_softmax, 1)
  115. loss = F.cross_entropy(y_hat_logits, y, weight=self.class_weights.to(self.device))
  117. # logging
  118. self.train_accuracy(predicted, y)
  120. self.log('train_loss', loss, batch_size = self.batch_size)
  121. self.log('train_accuracy', self.train_accuracy, on_step=True, prog_bar=True, on_epoch=True, batch_size = self.batch_size)
  123. return { 'loss': loss, 'preds': predicted, 'targets': y }
  125. def validation_step(self, batch, batch_idx):
  126. x, y, y_str, f_path = batch
  127. output = self(x)
  128. y_hat_logits = output[0]
  129. y_hat_softmax = output[1]
  130. _, predicted = torch.max(y_hat_softmax, 1)
  132. val_loss = F.cross_entropy(y_hat_logits, y, weight=self.class_weights.to(self.device))
  134. # logging
  135. self.val_accuracy(predicted, y)
  137. self.log('val_loss', val_loss, batch_size = self.batch_size)
  138. self.log('val_accuracy', self.val_accuracy, on_step=False, prog_bar=True, on_epoch=True, batch_size = self.batch_size)
  140. return { 'val_loss': val_loss, 'preds': predicted, 'targets': y }
  149. class b2v_FFwdNN_decrease_layer_sizes(pl.LightningModule):
  150. def __init__(self, input_dim, n_layers: int = 4, layer_dim: int = 1024, embedding_dim: int = 1024, dropout_rate: float = 0.0, nclasses: int = 2,class_weights: list = [1,1], batch_size: int = 16, lr: float = 0.001):
  151. super().__init__()
  152. self.save_hyperparameters()
  154. self.batch_size = batch_size # can be tuned by pytorchlightning
  155. self.lr = lr # can be tuned by pytorchlightning
  157. self.n_layers = n_layers
  159. self.dropout_rate = dropout_rate
  161. self.input_dim = input_dim
  162. self.layer_dim = layer_dim
  165. self.embedding_dim = embedding_dim
  166. self.nclasses = nclasses
  167. self.class_weights = class_weights
  170. self.base_net = torch.nn.Sequential()
  171. self.base_net.add_module('layer1_linear', nn.Linear(self.input_dim, self.layer_dim))
  172. self.base_net.add_module('layer1_relu', nn.LeakyReLU(inplace=True))
  173. self.base_net.add_module('layer1_batchnorm', nn.BatchNorm1d(self.layer_dim, momentum=0.1,affine=False))
  174. self.base_net.add_module('layer1_dropout',nn.Dropout(p=self.dropout_rate))
  176. self.prev_l_size = self.layer_dim
  177. self.next_l_size = int( self.prev_l_size / 2 )
  179. if self.n_layers > 1:
  180. for ix, ll in enumerate(list(range(1,self.n_layers))):
  181. self.base_net.add_module('layer'+ str(ll+1) + '_linear', nn.Linear(self.prev_l_size, self.next_l_size))
  182. self.base_net.add_module('layer'+ str(ll+1) + '_relu', nn.LeakyReLU(inplace=True))
  183. self.base_net.add_module('layer'+ str(ll+1) + '_batchnorm', nn.BatchNorm1d(self.next_l_size, momentum=0.1,affine=False))
  184. self.base_net.add_module('layer'+ str(ll+1) + '_dropout',nn.Dropout(p=self.dropout_rate))
  186. self.prev_l_size = self.next_l_size
  187. self.next_l_size = int( self.prev_l_size / 2 )
  189. self.embed1_linear = nn.Linear( self.prev_l_size ,self.embedding_dim)
  190. self.embed1_relu = nn.LeakyReLU(inplace=True)
  191. self.embed1_batchnorm = nn.BatchNorm1d(self.embedding_dim, momentum=0.1,affine=False)
  193. self.output = nn.Linear(self.embedding_dim, self.nclasses)
  194. init_weight(self)
  196. self.softmax = nn.Softmax(dim=1);
  199. # Metrics
  200. self.train_accuracy = torchmetrics.Accuracy(num_classes=self.nclasses, task = "multiclass")
  202. self.val_accuracy = torchmetrics.Accuracy(num_classes=self.nclasses, task = "multiclass")
  205. def forward(self, x):
  207. base_out = self.base_net.forward(x.float())
  209. x = self.embed1_linear(base_out)
  210. x = self.embed1_relu(x)
  211. embed = self.embed1_batchnorm(x)
  213. out_logit = self.output(embed)
  215. out_softmax = self.softmax(out_logit)
  217. return out_logit, out_softmax, embed, base_out
  219. def configure_optimizers(self):
  220. optimizer = optim.Adam(self.parameters(), lr=self.lr, weight_decay=1e-4)
  221. lr_schedulers = {'scheduler': torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode='min',patience=3,verbose=True),
  222. 'monitor': 'val_loss'}
  223. return [optimizer], [lr_schedulers]
  225. def training_step(self, batch, batch_idx):
  226. x, y, y_str, f_path = batch
  227. output = self(x)
  228. y_hat_logits = output[0]
  229. y_hat_softmax = output[1]
  230. _, predicted = torch.max(y_hat_softmax, 1)
  232. loss = F.cross_entropy(y_hat_logits, y, weight=self.class_weights.to(self.device))
  234. # logging
  235. self.train_accuracy(predicted, y)
  237. self.log('train_loss', loss, batch_size = self.batch_size)
  238. self.log('train_accuracy', self.train_accuracy, on_step=True, prog_bar=True, on_epoch=True, batch_size = self.batch_size)
  240. return { 'loss': loss, 'preds': predicted, 'targets': y }
  243. def validation_step(self, batch, batch_idx):
  244. x, y, y_str, f_path = batch
  245. output = self(x)
  246. y_hat_logits = output[0]
  247. y_hat_softmax = output[1]
  248. _, predicted = torch.max(y_hat_softmax, 1)
  250. val_loss = F.cross_entropy(y_hat_logits, y, weight=self.class_weights.to(self.device))
  252. # logging
  253. self.val_accuracy(predicted, y)
  255. self.log('val_loss', val_loss, batch_size = self.batch_size)
  256. self.log('val_accuracy', self.val_accuracy, on_step=False, prog_bar=True, on_epoch=True, batch_size = self.batch_size)
  258. return { 'val_loss': val_loss, 'preds': predicted, 'targets': y }
  264. class b2v_FFwdNN_embed_only(pl.LightningModule):
  265. def __init__(self, input_dim, embedding_dim: int = 1024, nclasses: int = 2, dropout_rate: float = 0.0,class_weights: list = [1,1], batch_size: int = 16, lr: float = 0.001):
  266. super().__init__()
  267. self.save_hyperparameters()
  269. self.batch_size = batch_size # can be tuned by pytorchlightning
  270. self.lr = lr # can be tuned by pytorchlightning
  272. self.dropout_rate = dropout_rate
  274. self.input_dim = input_dim
  276. self.embedding_dim = embedding_dim
  277. self.nclasses = nclasses
  278. self.class_weights = class_weights
  280. self.embed1_linear = nn.Linear(self.input_dim ,self.embedding_dim)
  281. self.embed1_relu = nn.LeakyReLU(inplace=True)
  282. self.embed1_batchnorm = nn.BatchNorm1d(self.embedding_dim, momentum=0.1,affine=False)
  283. self.embed1_do = nn.Dropout(p=self.dropout_rate)
  286. self.output = nn.Linear(self.embedding_dim, self.nclasses)
  287. init_weight(self)
  289. self.softmax = nn.Softmax(dim=1);
  292. # Metrics
  293. self.train_accuracy = torchmetrics.Accuracy(num_classes=self.nclasses, task = "multiclass")
  295. self.val_accuracy = torchmetrics.Accuracy(num_classes=self.nclasses, task = "multiclass")
  298. def forward(self, x):
  301. x = self.embed1_linear(x.float())
  302. x = self.embed1_relu(x)
  303. x = self.embed1_batchnorm(x)
  304. embed = x
  305. x = self.embed1_do(x)
  307. out_logit = self.output(x)
  309. out_softmax = self.softmax(out_logit)
  311. return out_logit, out_softmax, embed
  313. def configure_optimizers(self):
  314. optimizer = optim.Adam(self.parameters(), lr=self.lr, weight_decay=1e-4)
  315. lr_schedulers = {'scheduler': torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, mode='min',patience=3,verbose=True),
  316. 'monitor': 'val_loss'}
  317. return [optimizer], [lr_schedulers]
  319. def training_step(self, batch, batch_idx):
  320. x, y, y_str, f_path = batch
  321. output = self(x)
  322. y_hat_logits = output[0]
  323. y_hat_softmax = output[1]
  324. _, predicted = torch.max(y_hat_softmax, 1)
  326. loss = F.cross_entropy(y_hat_logits, y, weight=self.class_weights.to(self.device))
  328. # logging
  329. self.train_accuracy(predicted, y)
  331. self.log('train_loss', loss, batch_size = self.batch_size)
  332. self.log('train_accuracy', self.train_accuracy, on_step=True, prog_bar=True, on_epoch=True, batch_size = self.batch_size)
  334. return { 'loss': loss, 'preds': predicted, 'targets': y }
  336. def validation_step(self, batch, batch_idx):
  337. x, y, y_str, f_path = batch
  338. output = self(x)
  339. y_hat_logits = output[0]
  340. y_hat_softmax = output[1]
  341. _, predicted = torch.max(y_hat_softmax, 1)
  343. val_loss = F.cross_entropy(y_hat_logits, y, weight=self.class_weights.to(self.device))
  345. # logging
  346. self.val_accuracy(predicted, y)
  348. self.log('val_loss', val_loss, batch_size = self.batch_size)
  349. self.log('val_accuracy', self.val_accuracy, on_step=False, prog_bar=True, on_epoch=True, batch_size = self.batch_size)
  351. return { 'val_loss': val_loss, 'preds': predicted, 'targets': y }

brain2vec_model_arch.py, no license · at the source

Overview

Authors: Mattson Ogg1, Lindsey Kitchell1
ORCID iDs: Mattson Ogg
  1. Research and Exploratory Development Department, Johns Hopkins Applied Physics Laboratory, Laurel, Maryland 20723
Institutions: Johns Hopkins University Applied Physics Laboratory (United States)
Journal: eNeuro, volume 13, issue 8, pages ENEURO.0370-25.2026
Dates: received 6 October 2025; accepted 8 July 2026; published online 21 August 2026; in print August 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1523/eneuro.0370-25.2026 · PMID 42527301 · PMCID PMC13505883 · OpenAlex W7171691733
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: fMRI (modality), human (organism), methods / tools (subfield)
Methods: Connectivity, Statistics, Machine learning, Preprocessing, fMRI & imaging
Keywords: deep learning, fMRI, functional connectome, transfer learning
MeSH: Brain*, Connectome*, Deep Learning*, Functional Neuroimaging*, Magnetic Resonance Imaging*, Female, Humans, Male (* major topic)
Journal subjects: Research Article: New Research, Novel Tools and Methods
Topic: Functional Brain Connectivity Studies (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: Eli Lilly and Company; Johns Hopkins University; National Institutes of Health (P30 NS09857781, P30-AG066444, UL1‐TR000448, R01EB009352, R01AG043434, P01AG026276, 1U54‐MH‐091657, P01 AG003991, P50 AG00561, TR000448, AG003991, AG066444, AG026276); McDonnell Center for Systems Neuroscience (R01 EB009352, P01 AG026276, P30 NS09857781, R01 AG043434, P01 AG003991, 1U54MH091657, P50 AG00561, UL1 TR000448); Avid Radiopharmaceuticals
Citations: cited by 1 paper (Europe PMC); 115 references in the paper

Abstract

Functional MRI (fMRI) currently supports a limited application space stemming from modest dataset sizes, large interindividual variability, and heterogeneity among scanning protocols. These constraints have made it difficult for fMRI researchers to take full advantage of modern deep-learning tools that have revolutionized other fields such as NLP, speech transcription, and image recognition. To help address these issues, we scaled up functional connectome fingerprinting as a neural network pretraining task, drawing inspiration from speaker recognition research, to learn a generalizable representation of brain function. This approach achieves strong performance for neural fingerprinting on a previously unseen scale, across multiple public fMRI datasets (individual recognition from held-out scan sessions, 93% on MPI-Leipzig, 94% on NKI-Rockland, 73% on OASIS-3, and 99% on HCP). Performance is maintained even when evaluation scan duration is truncated to <2 min. We show that this representation can also generalize to support accurate neural fingerprinting for completely new datasets and participants of either sex not used in training. Finally, we demonstrate that the representation learned by the network encodes features related to individual variability that partially transfers to new tasks. These results support the development of scalable transfer-learning approaches for future clinical and cognitive neuroimaging applications.

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 5 matches between paragraphs and lines of code.

OSF wfdkx

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Size: 2 files
Software Heritage: not checked
Found in: “Data Availability”
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: NumPy (2 files), pandas (2 files), PyTorch (2 files), PyTorch Lightning (1 file), Matplotlib (1 file), Nilearn (1 file), scikit-learn (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)
  • 27 September 2026: the link answers (HTTP 200)
2 files
At the source: doi.org/10.17605/OSF.IO/WFDKX

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;
  • 2 scripts, each with its path and the digest of its content;
  • 5 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 Availability

All data used in this report are publicly accessible via https://openneuro.org, https://db.humanconnectome.org, https://fcon_1000.projects.nitrc.org and https://central.xnat.org. The preprocessing tools used in this report are available from https://fmriprep.org and https://nilearn.github.io. Pretrained model weights along with demo preprocessing and inference scripts are available in Extended Data 2 (https://doi.org/10.1523/ENEURO.0370-25.2026.d2). Code and pretrained model weights are also available at https://doi.org/10.17605/OSF.IO/WFDKX.

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 2, 28 September 2026

  • Funding: added Eli Lilly and Company; Johns Hopkins University; National Institutes of Health: P30 NS09857781, P30-AG066444, UL1‐TR000448, R01EB009352, R01AG043434, P01AG026276, 1U54‐MH‐091657, P01 AG003991, P50 AG00561, TR000448, AG003991, AG066444, AG026276; McDonnell Center for Systems Neuroscience: R01 EB009352, P01 AG026276, P30 NS09857781, R01 AG043434, P01 AG003991, 1U54MH091657, P50 AG00561, UL1 TR000448; Avid Radiopharmaceuticals

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 2 authors, 4 keywords, 8 MeSH terms, 96 references.

Cite

This paper

Ogg, M., & Kitchell, L. (2026). Pretraining for Large-Scale Functional Connectome Fingerprinting Supports Generalization and Transfer Learning in Functional Neuroimaging. eNeuro, 13(8), ENEURO.0370-25.2026. https://doi.org/10.1523/eneuro.0370-25.2026

BibTeX

@article{ogg2026pretraining,
author = {Ogg, Mattson and Kitchell, Lindsey},
title = {{Pretraining for Large-Scale Functional Connectome Fingerprinting Supports Generalization and Transfer Learning in Functional Neuroimaging}},
journal = {eNeuro},
year = {2026},
month = aug,
volume = {13},
number = {8},
pages = {ENEURO.0370--25.2026},
publisher = {Society for Neuroscience},
issn = {2373-2822},
doi = {10.1523/eneuro.0370-25.2026},
url = {https://doi.org/10.1523/eneuro.0370-25.2026},
pmid = {42527301},
pmcid = {PMC13505883}
}

RIS

TY - JOUR
AU - Ogg, Mattson
AU - Kitchell, Lindsey
TI - Pretraining for Large-Scale Functional Connectome Fingerprinting Supports Generalization and Transfer Learning in Functional Neuroimaging
T2 - eNeuro
J2 - eNeuro
PY - 2026
DA - 2026/08/24
VL - 13
IS - 8
SP - ENEURO.0370
EP - 25.2026
SN - 2373-2822
PB - Society for Neuroscience
DO - 10.1523/eneuro.0370-25.2026
UR - https://doi.org/10.1523/eneuro.0370-25.2026
LA - en
ER -

CSL-JSON

{
"id": "10.1523/eneuro.0370-25.2026",
"type": "article-journal",
"title": "Pretraining for Large-Scale Functional Connectome Fingerprinting Supports Generalization and Transfer Learning in Functional Neuroimaging",
"container-title": "eNeuro",
"author": [
{
"family": "Ogg",
"given": "Mattson"
},
{
"family": "Kitchell",
"given": "Lindsey"
}
],
"container-title-short": "eNeuro",
"volume": "13",
"issue": "8",
"page": "ENEURO.0370-25.2026",
"DOI": "10.1523/eneuro.0370-25.2026",
"PMID": "42527301",
"PMCID": "PMC13505883",
"ISSN": "2373-2822",
"publisher": "Society for Neuroscience",
"URL": "https://doi.org/10.1523/eneuro.0370-25.2026",
"language": "en",
"issued": {
"date-parts": [
[
2026,
8,
24
]
]
}
}

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:10.1371/journal.pbio.3003684 [code]
The retrieval of previously learned motor memories is facilitated by the reinstatement of default mode network manifold structures.
Journal: PLoS biology
In common: Nilearn, scikit-learn, pandas, 2 other tools, fMRI, 16 references
[2] doi:10.1038/s41467-026-71428-6 [code]
Binding items to contexts through conjunctive neural representations with the method of loci.
Journal: Nature communications
In common: Nilearn, scikit-learn, pandas, 2 other tools, 14 references
[3] doi:10.1371/journal.pbio.3003755 [code]
Action information is integrated into entorhinal representations of conceptual space and is reflected in eye movements.
Journal: PLoS biology
In common: Nilearn, pandas, Matplotlib, 1 other tool, 13 references
[4] doi:10.21203/rs.3.rs-9326213/v1 [code]
Multi-task fMRI outperforms resting-state fMRI for revealing task-invariant organization of the human brain
Journal: Research Square (preprint)
In common: Nilearn, PyTorch, scikit-learn, 3 other tools, fMRI, 10 references
[5] doi:10.64898/2026.03.09.710558 [code]
Multi-task fMRI outperforms resting-state fMRI for revealing task-invariant organization of the human brain
Journal: bioRxiv (preprint)
In common: Nilearn, PyTorch, scikit-learn, 3 other tools, fMRI, 10 references
[6] doi:10.1162/netn.a.547 [code]
An evaluation of the efficacy of single-echo and multi-echo fMRI denoising strategies.
Journal: Network neuroscience (Cambridge, Mass.)
In common: Nilearn, scikit-learn, pandas, 2 other tools, fMRI, 11 references
[7] doi:10.7554/elife.103956 [code]
Human brain-wide activation of sleep rhythms.
Journal: eLife
In common: Nilearn, pandas, NumPy, 13 references
[8] doi:10.1038/s41597-026-07323-y [code]
A densely sampled fMRI dataset for investigating food valuation.
Journal: Scientific data
In common: fMRI, methods / tools, 14 references
[9] doi:10.1162/imag.a.1272 [code]
Building bridges between brain and behavior: An open-source toolbox for joint modeling with fMRI.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: fMRI, methods / tools, 14 references
[10] doi:10.7554/elife.107955
Multiple event segmentation mechanisms in the human brain.
Journal: eLife
In common: 15 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.