Bayesian Uncertainty-aware Deep Learning with noisy labels: Tackling annotation ambiguity in EEG seizure detection.
5 blocks of code here have a permanent identifier (oscr:b:…), listed with the matches below: citable for good, with a page of its own. Select a match and press Cite to copy it. What a block identifier is.
The 5 matches · 1 of them tie a paragraph to a whole file without given lines: a weak match, whose lines are not tinted
- [1] § Materials and methods › Implementation and evaluation ↔ python/train/train_nal.py, lines 17 to 153 · score 0.63 · validation loss, PyTorch, finetuning, Adam, NAL, optimizer · block
oscr:b:FSEZAWRHMF8NQ-A· Report a mismatch - [2] § Materials and methods › Implementation and evaluation ↔ python/train/train_hvit.py, lines 19 to 134 · score 0.62 · validation loss, PyTorch, HViT, Adam, DUL, optimizer · block
oscr:b:A8X2THCH317RV-C· Report a mismatch - [3] § Materials and methods › EEG datasets › Simulated data. ↔ matlab/simulate/SEREEGA-master/pop/pop_sereega_plot_source_location.m, lines 40 to 131 · score 0.53 · source locations, lead field, SEREEGA, components, simulate · block
oscr:b:6H9CSX03VJPST-D· Report a mismatch - [4] § Materials and methods › BUNDL: Bayesian Uncertainty-aware Deep Learning › Approximating label mismatch with . ↔ python/train/BUNDL.py, lines 73 to 156 · score 0.52 · cross entropy loss, ensemble, TTA, MCD, uncertainty, predictions · block
oscr:b:QDQSPKDGWPKE0-E· Report a mismatch - [5] § Materials and methods › EEG datasets › Simulated data. ↔ matlab/simulate/get_signal.m, the whole file · a weak match · score 0.51 · sharp waves, polyspike, spikes, Seizure · block
oscr:b:GPE5Q6TRHEERE-P· 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 · 153 lines · 6.2 KB · no license · 1 match
- from models_nal import *
- from utils import *
- #from dataloader import *
- import numpy as np
- import pandas as pd
- import json
- import os
- import scipy
- import scipy.special
- import torch
- import torch.nn as nn
- import sklearn.metrics as metrics
- #from pyt_enc import *
- import torch.nn.functional as F
- detloss = nn.CrossEntropyLoss(weight = torch.Tensor([0.2, 0.8]).double().to('cuda:0'))
- def nal_finetune(data_root, modelname, mn_fn, cvfold,
- hparams = {'lr':1e-04, 'maxiter':100, 'prior':0.8, 'beta':1.0},
- traintype='s_model',
- use_cuda=False):
- lr = hparams['lr']
- maxiter = hparams['maxiter']
- beta = hparams['beta']
- prior = hparams['prior']
- device = 'cuda:0' if use_cuda and torch.cuda.is_available() else 'cpu'
- if use_cuda:
- torch.cuda.empty_cache()
- print(traintype, modelname)
- #detloss = nn.CrossEntropyLoss(weight = torch.Tensor([0.2, 0.8]).double().to(device))
- manifest = read_manifest(data_root+mn_fn, ',')
- save_loc = data_root+'split'+str(cvfold)+'/models/'
- if not os.path.exists(save_loc):
- os.mkdir(save_loc)
- pt_list = np.load(data_root+'all_pts.npy')
- train_set = myDataSet(data_root,cvfold, manifest,
- normalize=True,train=True,
- maxSeiz = 10)
- val_set = myDataSet(data_root,cvfold, manifest,
- normalize=True,train=False,
- maxSeiz = 10)
- train_loader = DataLoader(train_set, batch_size=1, shuffle=True)
- val_loader = DataLoader(val_set, batch_size=1, shuffle=True)
- s_model = True if traintype=='s_model' else False
- cm_theta = torch.tensor(np.load(modelname.split('_')[0]+str(cvfold)+'_cmtheta_pretrain_unc.npy')).double()
- if modelname == 'txlstm_nal':
- model = transformer_lstm_nal(transformer_dropout=0.2, device=device, s_model=s_model, cm_theta = cm_theta)
- elif modelname == 'sztrack_nal':
- model = sztrack_nal( s_model=s_model, cm_theta = cm_theta)
- elif modelname == 'tgcn_nal':
- model = TGCN_nal(s_model=s_model, cm_theta = cm_theta, device=device)
- elif modelname == 'tx3lstm_nal':
- model = transformer_3lstm(transformer_dropout=0.1, device=device)
- else:
- model = CNN_BLSTM_nal(s_model=s_model, cm_theta = cm_theta, device=device)
- savename = modelname+'_'+traintype+'_cv'+str(cvfold)+'_'+str(lr)
- oldlr = {'txlstm':1e-05, 'tgcn': 1e-06, 'cnnlstm':0.01, 'cnnblstm':0.01}
- ########## initialize with pretrained model, confusion matrix of pretrained or s_model
- if s_model:
- oldsavename = modelname.split('_')[0]+'_pretrain_unc_cv'+str(cvfold)+'_'+str(oldlr[modelname.split('_')[0]])
- state_dict = torch.load(save_loc + oldsavename+'.pth.tar')
- with torch.no_grad():
- for n, p in model.named_parameters():
- try:
- p.data = state_dict[n].double()
- except:
- continue
- else: #c_model case
- oldsavename = modelname+'_s_model_cv'+str(cvfold)+'_'+str(oldlr[modelname.split('_')[0]])
- state_dict = torch.load(save_loc + oldsavename+'.pth.tar')
- #for n, p in state_dict.named_parameters():
- # model[n].copy_(p)
- model.load_state_dict(state_dict)
- #print(nothing)
- model.double()
- model.to(device)
- optimizer = torch.optim.Adam(model.parameters(), lr=lr)
- sigmoid = torch.nn.Sigmoid()
- train_loss_clean = []
- train_loss_corrupt = []
- val_losses = []
- train_losses = []
- update_labels = False
- for epoch in range(1, maxiter+1):
- epoch_loss = 0.0
- train_len = len(train_loader)
- val_len = len(val_loader)
- val_epoch_loss = 0.0
- for batch_idx, data in enumerate(train_loader):
- optimizer.zero_grad()
- inputs = data['buffers']
- det_labels = data['sz_labels'].long().to(device)
- b, Nsz, T, _, _ = inputs.shape
- inputs = inputs.to(torch.DoubleTensor()).to(device)
- k_pred, noise_pred, _ = model(inputs)
- y_samples = []
- loss = 0.5* detloss(k_pred.reshape(-1, 2), det_labels.reshape(-1)) + 0.5* detloss(noise_pred.reshape(-1, 2), det_labels.reshape(-1))
- loss.backward()
- optimizer.step()
- epoch_loss += loss.item()
- for batch_idx, data in enumerate(val_loader):
- with torch.no_grad():
- optimizer.zero_grad()
- inputs = data['buffers']
- det_labels = data['sz_labels'].long().to(device)
- b, Nsz, T, _, _ = inputs.shape
- inputs = inputs.to(torch.DoubleTensor()).to(device)
- k_pred, noise_pred, _ = model(inputs)
- y_samples = []
- loss = 0.5* detloss(k_pred.reshape(-1, 2), det_labels.reshape(-1)) + 0.5* detloss(noise_pred.reshape(-1, 2), det_labels.reshape(-1))
- val_epoch_loss += loss.item()
- epoch_loss = epoch_loss/train_len
- val_epoch_loss = val_epoch_loss/val_len
- torch.cuda.empty_cache()
- #schedule lr
- train_losses.append(epoch_loss)
- val_losses.append(val_epoch_loss)
- if len(val_losses) > 10 and val_losses[-1] > min(val_losses[-11:-1]):
- break
- #if epoch_loss <= train_losses[-1] and epoch_val_loss <= val_losses[-1]:
- #torch.save(model.state_dict(), '/home/deeksha/EEG_Sz/GenProc/results/lstmAE_'+pt+str(ch_id)+'.pth.tar')
- print('Epoch: {} \tTraining Loss: {:.6f} \tValidation Loss: {:.6f} '.format(epoch, epoch_loss, val_epoch_loss))
- torch.save(model.state_dict(), save_loc+savename+ '.pth.tar')
- np.save(save_loc+ savename+'_train.npy', train_losses)
- np.save(save_loc+savename+'_val.npy', val_losses)
- #del model, optimizer, train_loader, validation_loader
- return train_losses, val_losses
train_nal.py at commit 3029ac2, no license · at the source
Overview
- Department of Electrical and Computer Engineering, Johns Hopkins University, Baltimore, Maryland, United States of America
- Department of Electrical and Computer Engineering, Boston University, Boston, Massachusetts, United States of America
Abstract
Deep learning is advancing EEG processing for automated epileptic seizure detection and onset zone localization, yet its performance relies heavily on high-quality annotated training data. However, scalp EEG is susceptible to high noise levels, which in turn leads to imprecise annotations of the seizure timing and characteristics. This “label noise” presents a significant challenge in model training and generalization. In this paper, we introduce Bayesian UncertaiNty-aware Deep Learning (BUNDL), a novel algorithm that informs a deep learning model of label ambiguities, thereby enhancing the robustness of seizure detection systems. By integrating domain knowledge into an underlying Bayesian framework, we derive a novel KL-divergence-based loss function that capitalizes on uncertainty to better learn seizure characteristics from scalp EEG. Thus, BUNDL offers a straightforward and model-agnostic method for training deep neural networks with noisy training labels that does not add any parameters to existing architectures. Additionally, we explore the impact of improved detection system on the task of automated onset zone localization. We validate BUNDL using a comprehensive simulated EEG dataset and two publicly available datasets collected by Temple University Hospital (TUH) and Boston Children’s Hospital (CHB-MIT). Results show that BUNDL consistently identifies noisy labels and improves the robustness of three base models under various label noise conditions. We also conduct ablation experiments on uncertainty quantification, evaluate cross-site generalizability to Siena EEG dataset, and quantify computational cost of all methods. Furthermore, we demonstrate that BUNDL improves seizure onset zone localization accuracy. Ultimately, BUNDL presents as a reliable method that can be seamlessly integrated with existing deep models used in clinical practice, enabling the training of trustworthy models for epilepsy evaluation.
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.
deeksha-ms/BUNDL
3029ac2e1a5d3dfd88529c459e6dca93566c9f02, 3 January 2026Availability: 1 check, the latest on 27 September 2026: the link answers
- 27 September 2026: the link answers
144 files
- matlab/
simulate/ , MATLAB, 166 linesSEREEGA-master/ eegplugin_sereega.m - matlab/
simulate/ , MATLAB, 220 linesSEREEGA-master/ generate/ generate_scalpdata.m - matlab/
simulate/ , MATLAB, 76 linesSEREEGA-master/ generate/ generate_signal_fromclas s.m - matlab/
simulate/ , MATLAB, 79 linesSEREEGA-master/ generate/ generate_signal_fromcomp onent.m - matlab/
simulate/ , MATLAB, 467 linesSEREEGA-master/ leadfield/ brainstorm/ lf_generate_frombrainsto rm.m - matlab/
simulate/ , MATLAB, 173 linesSEREEGA-master/ leadfield/ fieldtrip/ lf_generate_fromfieldtri p.m - matlab/
simulate/ , MATLAB, 144 linesSEREEGA-master/ leadfield/ hartmut/ lf_generate_fromhartmut. m - matlab/
simulate/ , MATLAB, 154 linesSEREEGA-master/ leadfield/ lf_add_atlas_frommni2atl as.m - matlab/
simulate/ , MATLAB, 109 linesSEREEGA-master/ leadfield/ lf_add_source.m - matlab/
simulate/ , MATLAB, 123 linesSEREEGA-master/ leadfield/ lf_get_projection.m - matlab/
simulate/ , MATLAB, 75 linesSEREEGA-master/ leadfield/ lf_get_regions.m - matlab/
simulate/ , MATLAB, 82 linesSEREEGA-master/ leadfield/ lf_get_source_all.m - matlab/
simulate/ , MATLAB, 79 linesSEREEGA-master/ leadfield/ lf_get_source_inradius.m - matlab/
simulate/ , MATLAB, 80 linesSEREEGA-master/ leadfield/ lf_get_source_middle.m - matlab/
simulate/ , MATLAB, 77 linesSEREEGA-master/ leadfield/ lf_get_source_nearest.m - matlab/
simulate/ , MATLAB, 82 linesSEREEGA-master/ leadfield/ lf_get_source_random.m - matlab/
simulate/ , MATLAB, 139 linesSEREEGA-master/ leadfield/ lf_get_source_spaced.m - matlab/
simulate/ , MATLAB, 94 linesSEREEGA-master/ leadfield/ lf_project_signal.m - matlab/
simulate/ , MATLAB, 128 linesSEREEGA-master/ leadfield/ nyhead/ lf_generate_fromnyhead.m - matlab/
simulate/ , MATLAB, 81 linesSEREEGA-master/ leadfield/ nyhead/ lf_prune_nyheadfile.m - matlab/
simulate/ , MATLAB, 205 linesSEREEGA-master/ leadfield/ pha/ lf_fix_electrodes_pha.m - matlab/
simulate/ , MATLAB, 202 linesSEREEGA-master/ leadfield/ pha/ lf_generate_frompha.m - matlab/
simulate/ , MATLAB, 95 linesSEREEGA-master/ plot/ plot_chanlocs_dipplot.m - matlab/
simulate/ , MATLAB, 68 linesSEREEGA-master/ plot/ plot_component.m - matlab/
simulate/ , MATLAB, 92 linesSEREEGA-master/ plot/ plot_component_projectio n.m - matlab/
simulate/ , MATLAB, 89 linesSEREEGA-master/ plot/ plot_component_signal.m - matlab/
simulate/ , MATLAB, 140 linesSEREEGA-master/ plot/ plot_headmodel.m - matlab/
simulate/ , MATLAB, 76 linesSEREEGA-master/ plot/ plot_signal_fromclass.m - matlab/
simulate/ , MATLAB, 180 linesSEREEGA-master/ plot/ plot_source_location.m - matlab/
simulate/ , MATLAB, 136 linesSEREEGA-master/ plot/ plot_source_location_dip plot.m - matlab/
simulate/ , MATLAB, 80 linesSEREEGA-master/ plot/ plot_source_moment.m - matlab/
simulate/ , MATLAB, 146 linesSEREEGA-master/ plot/ plot_source_projection.m - matlab/
simulate/ , MATLAB, 136 linesSEREEGA-master/ pop/ pop_sereega_addarm.m - matlab/
simulate/ , MATLAB, 101 linesSEREEGA-master/ pop/ pop_sereega_adddata.m - matlab/
simulate/ , MATLAB, 115 linesSEREEGA-master/ pop/ pop_sereega_adderp.m - matlab/
simulate/ , MATLAB, 228 linesSEREEGA-master/ pop/ pop_sereega_addersp.m - matlab/
simulate/ , MATLAB, 93 linesSEREEGA-master/ pop/ pop_sereega_addnoise.m - matlab/
simulate/ , MATLAB, 206 linesSEREEGA-master/ pop/ pop_sereega_components.m - matlab/
simulate/ , MATLAB, 88 linesSEREEGA-master/ pop/ pop_sereega_epochs.m - matlab/
simulate/ , MATLAB, 114 linesSEREEGA-master/ pop/ pop_sereega_lf_generate_ fromfieldtrip.m - matlab/
simulate/ , MATLAB, 149 linesSEREEGA-master/ pop/ pop_sereega_lf_generate_ fromhartmut.m - matlab/
simulate/ , MATLAB, 111 linesSEREEGA-master/ pop/ pop_sereega_lf_generate_ fromnyhead.m - matlab/
simulate/ , MATLAB, 157 linesSEREEGA-master/ pop/ pop_sereega_lf_generate_ frompha.m - matlab/
simulate/ , MATLAB, 93 linesSEREEGA-master/ pop/ pop_sereega_lftoggle.m - matlab/
simulate/ , MATLAB, 117 linesSEREEGA-master/ pop/ pop_sereega_plot_headmod el.m - matlab/
simulate/ , MATLAB, 131 lines, 1 matchSEREEGA-master/ pop/ pop_sereega_plot_source_ location.m - matlab/
simulate/ , MATLAB, 186 linesSEREEGA-master/ pop/ pop_sereega_signals.m - matlab/
simulate/ , MATLAB, 514 linesSEREEGA-master/ pop/ pop_sereega_sources.m - matlab/
simulate/ , MATLAB, 100 linesSEREEGA-master/ pop/ pop_sereega_utl_mix_data .m - matlab/
simulate/ , MATLAB, 85 linesSEREEGA-master/ pop/ pop_sereega_utl_reorder_ eeglabdataset.m - matlab/
simulate/ , MATLAB, 95 linesSEREEGA-master/ signal/ arm/ arm_check_class.m - matlab/
simulate/ , MATLAB, 41 linesSEREEGA-master/ signal/ arm/ arm_class2string.m - matlab/
simulate/ , MATLAB, 120 linesSEREEGA-master/ signal/ arm/ arm_generate_signal.m - matlab/
simulate/ , MATLAB, 87 linesSEREEGA-master/ signal/ arm/ arm_generate_signal_from class.m - matlab/
simulate/ , MATLAB, 149 linesSEREEGA-master/ signal/ arm/ arm_get_class_interactin g.m - matlab/
simulate/ , MATLAB, 90 linesSEREEGA-master/ signal/ arm/ arm_plot_signal_fromclas s.m - matlab/
simulate/ , MATLAB, 116 linesSEREEGA-master/ signal/ data/ data_check_class.m - matlab/
simulate/ , MATLAB, 44 linesSEREEGA-master/ signal/ data/ data_class2string.m - matlab/
simulate/ , MATLAB, 113 linesSEREEGA-master/ signal/ data/ data_generate_signal_fro mclass.m - matlab/
simulate/ , MATLAB, 81 linesSEREEGA-master/ signal/ data/ data_plot_signal_fromcla ss.m - matlab/
simulate/ , MATLAB, 160 linesSEREEGA-master/ signal/ erp/ erp_check_class.m - matlab/
simulate/ , MATLAB, 45 linesSEREEGA-master/ signal/ erp/ erp_class2string.m - matlab/
simulate/ , MATLAB, 76 linesSEREEGA-master/ signal/ erp/ erp_generate_signal.m - matlab/
simulate/ , MATLAB, 90 linesSEREEGA-master/ signal/ erp/ erp_generate_signal_from class.m - matlab/
simulate/ , MATLAB, 121 linesSEREEGA-master/ signal/ erp/ erp_get_class_random.m - matlab/
simulate/ , MATLAB, 134 linesSEREEGA-master/ signal/ erp/ erp_plot_signal_fromclas s.m - matlab/
simulate/ , MATLAB, 334 linesSEREEGA-master/ signal/ ersp/ ersp_check_class.m - matlab/
simulate/ , MATLAB, 48 linesSEREEGA-master/ signal/ ersp/ ersp_class2string.m - matlab/
simulate/ , MATLAB, 239 linesSEREEGA-master/ signal/ ersp/ ersp_generate_signal.m - matlab/
simulate/ , MATLAB, 145 linesSEREEGA-master/ signal/ ersp/ ersp_generate_signal_fro mclass.m - matlab/
simulate/ , MATLAB, 178 linesSEREEGA-master/ signal/ ersp/ ersp_get_class_randomamp mod.m - matlab/
simulate/ , MATLAB, 195 linesSEREEGA-master/ signal/ ersp/ ersp_get_class_randombur st.m - matlab/
simulate/ , MATLAB, 136 linesSEREEGA-master/ signal/ ersp/ ersp_get_class_randomnom od.m - matlab/
simulate/ , MATLAB, 84 linesSEREEGA-master/ signal/ ersp/ ersp_plot_signal_fromcla ss.m - matlab/
simulate/ , MATLAB, 120 linesSEREEGA-master/ signal/ noise/ noise_check_class.m - matlab/
simulate/ , MATLAB, 41 linesSEREEGA-master/ signal/ noise/ noise_class2string.m - matlab/
simulate/ , MATLAB, 112 linesSEREEGA-master/ signal/ noise/ noise_generate_signal_co loreduniform.m - matlab/
simulate/ , MATLAB, 132 linesSEREEGA-master/ signal/ noise/ noise_generate_signal_fr omclass.m - matlab/
simulate/ , MATLAB, 115 linesSEREEGA-master/ signal/ noise/ noise_plot_signal_fromcl ass.m - matlab/
simulate/ , MATLAB, 54 linesSEREEGA-master/ utils/ utl_add_drift.m - matlab/
simulate/ , MATLAB, 45 linesSEREEGA-master/ utils/ utl_add_icaweights_toeeg labdataset.m - matlab/
simulate/ , MATLAB, 66 linesSEREEGA-master/ utils/ utl_add_markers_toeeglab dataset.m - matlab/
simulate/ , MATLAB, 58 linesSEREEGA-master/ utils/ utl_add_sensornoise.m - matlab/
simulate/ , MATLAB, 95 linesSEREEGA-master/ utils/ utl_add_signal_tocompone nt.m - matlab/
simulate/ , MATLAB, 58 linesSEREEGA-master/ utils/ utl_apply_dvslope.m - matlab/
simulate/ , MATLAB, 79 linesSEREEGA-master/ utils/ utl_apply_dvslopeshift.m - matlab/
simulate/ , MATLAB, 66 linesSEREEGA-master/ utils/ utl_call_gui.m - matlab/
simulate/ , MATLAB, 73 linesSEREEGA-master/ utils/ utl_check_class.m - matlab/
simulate/ , MATLAB, 102 linesSEREEGA-master/ utils/ utl_check_component.m - matlab/
simulate/ , MATLAB, 49 linesSEREEGA-master/ utils/ utl_class2string.m - matlab/
simulate/ , MATLAB, 84 linesSEREEGA-master/ utils/ utl_create_component.m - matlab/
simulate/ , MATLAB, 114 linesSEREEGA-master/ utils/ utl_create_eeglabdataset .m - matlab/
simulate/ , MATLAB, 134 linesSEREEGA-master/ utils/ utl_epoch2continuous.m - matlab/
simulate/ , MATLAB, 361 linesSEREEGA-master/ utils/ utl_get_component_fromte mplate.m - matlab/
simulate/ , MATLAB, 52 linesSEREEGA-master/ utils/ utl_get_icaweights.m - matlab/
simulate/ , MATLAB, 109 linesSEREEGA-master/ utils/ utl_get_montage.m - matlab/
simulate/ , MATLAB, 51 linesSEREEGA-master/ utils/ utl_get_orientation_pseu doperpendicular.m - matlab/
simulate/ , MATLAB, 56 linesSEREEGA-master/ utils/ utl_get_orientation_pseu dotangential.m - matlab/
simulate/ , MATLAB, 47 linesSEREEGA-master/ utils/ utl_get_orientation_rand om.m - matlab/
simulate/ , MATLAB, 48 linesSEREEGA-master/ utils/ utl_iscomponent.m - matlab/
simulate/ , MATLAB, 53 linesSEREEGA-master/ utils/ utl_isvalidcomponent.m - matlab/
simulate/ , MATLAB, 93 linesSEREEGA-master/ utils/ utl_merge_components.m - matlab/
simulate/ , MATLAB, 95 linesSEREEGA-master/ utils/ utl_mix_data.m - matlab/
simulate/ , MATLAB, 624 linesSEREEGA-master/ utils/ utl_multigradient.m - matlab/
simulate/ , MATLAB, 49 linesSEREEGA-master/ utils/ utl_normalise.m - matlab/
simulate/ , MATLAB, 41 linesSEREEGA-master/ utils/ utl_randsample.m - matlab/
simulate/ , MATLAB, 88 linesSEREEGA-master/ utils/ utl_reorder_eeglabdatase t.m - matlab/
simulate/ , MATLAB, 67 linesSEREEGA-master/ utils/ utl_sanitize_atlas.m - matlab/
simulate/ , MATLAB, 92 linesSEREEGA-master/ utils/ utl_set_dvslope.m - matlab/
simulate/ , MATLAB, 80 linesSEREEGA-master/ utils/ utl_shift_latency.m - matlab/
simulate/ , MATLAB, 65 linesSEREEGA-master/ utils/ utl_shift_source.m - matlab/
simulate/ , MATLAB, 98 linesSEREEGA-master/ utils/ utl_show_progress_parfor .m - matlab/
simulate/ , MATLAB, 163 linesSEREEGA-master/ utils/ utl_show_progress_timeto complete.m - matlab/
simulate/ , MATLAB, 22 linesgenerate_background.m - matlab/
simulate/ , MATLAB, 35 linesgenerate_gauss.m - matlab/
simulate/ , MATLAB, 49 linesgenerate_seizure.m - matlab/
simulate/ , MATLAB, 21 linesgenerate_slow.m - matlab/
simulate/ , MATLAB, 23 linesgenerate_spikes.m - matlab/
simulate/ , MATLAB, 48 linesgenereate_seizure.m - matlab/
simulate/ , MATLAB, 23 linesget_eye_artifact.m - matlab/
simulate/ , MATLAB, 44 linesget_musc_artifact.m - matlab/
simulate/ , MATLAB, 56 linesget_noise.m - matlab/
simulate/ , MATLAB, 25 lines, 1 matchget_signal.m - matlab/
simulate/ , MATLAB, 27 linesoverestimate.m - matlab/
simulate/ , MATLAB, to read at its sourcesave_noisylabel.mlx - matlab/
simulate/ , MATLAB, to read at its sourcesimulate_dummy.mlx - matlab/
simulate/ , MATLAB, to read at its sourcesimulate_signal_bits.mlx - matlab/
simulate/ , MATLAB, to read at its sourcesimulate_v1.mlx - matlab/
simulate/ , MATLAB, to read at its sourcesimulate_v2.mlx - matlab/
simulate/ , MATLAB, to read at its sourcetutorial.mlx - matlab/
simulate/ , MATLAB, 20 linesundersample.m - python/
preprocess/ , Jupyter, 778 linespreprocess_siena.ipynb - python/
preprocess/ , Jupyter, 210 linespreprocess_tuh.ipynb - python/
train/ , Python, 157 lines, 1 matchBUNDL.py - python/
train/ , Python, 181 linesdataloader.py - python/
train/ , Python, 153 lineshvit.py - python/
train/ , Python, 273 linesmodels.py - python/
train/ , Python, 309 linesmodels_nal.py - python/
train/ , Python, 224 linestrain_bundl.py - python/
train/ , Python, 122 linestrain_cel.py - python/
train/ , Python, 268 lines, 1 matchtrain_hvit.py - python/
train/ , Python, 153 lines, 1 matchtrain_nal.py - python/
train/ , Python, 146 linestrain_selfadapt.py - README.md, Text, 14 lines
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;
- 143 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
Datasets cited
- physionet.org/
content/ , at PhysioNet; found in “Data Availability”chbmit - physionet.org/
content/ , at PhysioNet; found in “Data Availability”siena-scalp-eeg
Data Availability
The simulated data can be reproduced using scripts available at https://
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 National Science Foundation: CAREER, 1845430; National Institutes of Health: R21 CA263804, R01 EB029977, CA263804, R01-HD108790, EB029977
Version 1, 27 September 2026: the first record
Recorded: type, language, journal, volume, issue, pages, dates, 2 authors, 8 MeSH terms, 44 references.
Cite
This paper
M Shama, D., & Venkataraman, A. (2026). Bayesian Uncertainty-aware Deep Learning with noisy labels: Tackling annotation ambiguity in EEG seizure detection. PloS one, 21(6), e0352191. https://
BibTeX
@article{mshama2026bayes
author = {M Shama, Deeksha and Venkataraman, Archana},
title = {{Bayesian Uncertainty-aware Deep Learning with noisy labels: Tackling annotation ambiguity in EEG seizure detection}},
journal = {PloS one},
year = {2026},
month = jun,
volume = {21},
number = {6},
pages = {e0352191},
publisher = {PLOS},
issn = {1932-6203},
doi = {10.1371/
url = {https://
pmid = {42335163},
pmcid = {PMC13289946}
}
RIS
TY - JOUR
AU - M Shama, Deeksha
AU - Venkataraman, Archana
TI - Bayesian Uncertainty-aware Deep Learning with noisy labels: Tackling annotation ambiguity in EEG seizure detection
T2 - PloS one
J2 - PLoS One
PY - 2026
DA - 2026/
VL - 21
IS - 6
SP - e0352191
SN - 1932-6203
PB - PLOS
DO - 10.1371/
UR - https://
LA - en
ER -
CSL-JSON
{
"id": "10.1371/
"type": "article-journal",
"title": "Bayesian Uncertainty-aware Deep Learning with noisy labels: Tackling annotation ambiguity in EEG seizure detection",
"container-title": "PloS one",
"author": [
{
"family": "M Shama",
"given": "Deeksha"
},
{
"family": "Venkataraman",
"given": "Archana"
}
],
"container-title-short":
"volume": "21",
"issue": "6",
"page": "e0352191",
"DOI": "10.1371/
"PMID": "42335163",
"PMCID": "PMC13289946",
"ISSN": "1932-6203",
"publisher": "PLOS",
"URL": "https://
"language": "en",
"issued": {
"date-parts": [
[
2026,
6,
23
]
]
}
}
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:10.3389/fnins.2023.1191683
- Epileptic seizures detection and the analysis of optimal seizure prediction horizon based on frequency and phase analysis.Journal: Frontiers in neuroscienceIn common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, methods / tools, EEG, 3 references
- [2] doi:10.1186/s42492-025-00208-8
- Comprehensive review of machine learning and deep learning techniques for epileptic seizure detection and prediction based on neuroimaging modalities.Journal: Visual computing for industry, biomedicine, and artIn common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, EEG, 3 references
- [3] doi:10.3390/s26030863
- Explainable AI-Enhanced Ensemble Protocol Using Gradient-Boosted Models for Zero-False-Alarm Seizure Detection from EEG.Journal: Sensors (Basel, Switzerland)In common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, EEG, 3 references
- [4] doi:10.3389/fphys.2023.1227952
- Automatic epileptic seizure detection based on persistent homology.Journal: Frontiers in physiologyIn common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, methods / tools, EEG, 2 references
- [5] doi:10.1038/s41598-025-88492-5
- Synchronization-based graph spatio-temporal attention network for seizure prediction.Journal: Scientific reportsIn common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, EEG, 2 references
- [6] doi:10.3390/brainsci14040306
- Evaluation of the Relation between Ictal EEG Features and XAI Explanations.Journal: Brain sciencesIn common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, methods / tools, EEG, 2 references
- [7] doi:10.3390/s22083066
- Evaluation of Feature Selection Methods for Classification of Epileptic Seizure EEG Signals.Journal: Sensors (Basel, Switzerland)In common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, methods / tools, EEG, 2 references
- [8] doi:10.1098/rsos.230601 [code]
- Epileptic seizure detection using CHB-MIT dataset: The overlooked perspectives.Journal: Royal Society open scienceIn common: scikit-learn, pandas, SciPy, 2 other tools, physionet.org/content/chbmit, epilepsy, methods / tools, EEG, 2 references
- [9] doi:10.3390/biomedicines12061283
- Graphical Insight: Revolutionizing Seizure Detection with EEG Representation.Journal: BiomedicinesIn common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, EEG, 2 references
- [10] doi:10.1002/epi4.12704 [code]
- EEG datasets for seizure detection and prediction- A review.Journal: Epilepsia openIn common: physionet.org/content/siena-scalp-eeg, physionet.org/content/chbmit, epilepsy, methods / tools, EEG, 1 reference
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.
Claim this paper
Correct its record
Say what each link of this record is, remove the ones that are not the paper's, add the ones that are missing. The correction becomes a new version of the record, in its Versions section.
Validate its tracing map
You validate the map as this page shows it: 1 repository of the authors' code, each at its verified commit and with its license, 143 scripts, and 5 matches between paragraphs and code (see the Code and Map sections). It then receives a DOI on Zenodo, with you (your ORCID iD) and OSCR as its creators; the code itself is not deposited.
The map's fingerprint: sha256:2d4e2686a32a7caa…
Add the badge to the code's README
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 -->
[](https://openscicode.org/badge/doi/10.1371/journal.pone.0352191)
<!-- oscr-badge:end -->reStructuredText
.. oscr-badge:start
.. image:: https://openscicode.org/badge/doi/10.1371/journal.pone.0352191.svg
:target: https://openscicode.org/badge/doi/10.1371/journal.pone.0352191
:alt: OSCR
.. oscr-badge:endHTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.1371/journal.pone.0352191"><img src="https://openscicode.org/badge/doi/10.1371/journal.pone.0352191.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.
