OSCR

Validation of portable, semi-dry electrode-based electroencephalography device for its application in brain-computer interface solutions.

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.

4 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. [1] § Materials and methods › Machine learning methods › Convolutional neural networks ↔ Preprocessing/MI.py, lines 246 to 294 · score 0.82 · categorical crossentropy, Adam, monitoring, optimizer, patience, epochs · block oscr:b:0GP897PGSPSCJ-J · Report a mismatch
  2. [2] § Materials and methods › Machine learning methods › Support vector machines ↔ Preprocessing/MI.py, lines 187 to 244 · score 0.74 · Nu SVC, class weight, kernel, vector, SVM · block oscr:b:YFAV6M3N4C7ZB-D · Report a mismatch
  3. [3] § Materials and methods › Machine learning methods › Convolutional neural networks ↔ Preprocessing/MI.py, lines 246 to 294 · score 0.70 · Keras, padding, ReLU, activation, dense, dropout · block oscr:b:0GP897PGSPSCJ-J · Report a mismatch
  4. [4] § Materials and methods › BCI paradigms › Visually evoked potentials ↔ Preprocessing/VEP.py, lines 10 to 63 · score 0.55 · FastICA, SOS, components, threshold, 25 Hz, VEP · block oscr:b:VJJW6XRWT4V7S-X · Report a mismatch
  5. [5] § Materials and methods › BCI paradigms › Visually evoked potentials ↔ Preprocessing/MI.py, lines 12 to 63 · score 0.51 · FastICA, SOS, components, filtered · block oscr:b:B9Y2HXPFB1M2Y-A · 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 · 294 lines · 13 KB · no license · 4 matches

  1. import numpy as np
  2. import matplotlib.pyplot as plt
  3. from scipy import signal
  4. from sklearn.decomposition import FastICA
  5. from scipy.fft import fft
  6. from sklearn import svm
  7. from sklearn.ensemble import RandomForestClassifier
  8. from sklearn.metrics import confusion_matrix, accuracy_score
  9. import keras
  10. from keras import layers
  11. #ifn = "MI_VSN_2024-08-01_13-09-52.csv"
  12. ifn = "MI_Smarting_2024-07-31_14-28-05.csv"
  13. fs = 250
  14. data_str = np.loadtxt(ifn, dtype=str, delimiter=';')
  15. data = np.zeros(data_str.shape)
  16. if ifn.split(sep='_')[1] == "VSN":
  17. data = data_str[:, 0:8].astype(np.float32) * 0.045
  18. data = np.delete(data, obj=6, axis=1)
  19. else:
  20. for i in range(data.shape[0]):
  21. for j in range(data.shape[1]):
  22. data[i, j] = float(data_str[i, j].replace(',', '.'))
  23. data = data[:, 0:-2]
  24. ref = data[:, 5]
  25. for i in range(data.shape[1]):
  26. data[:, i] = data[:, i] - ref
  27. data = np.delete(data, obj=[5, 6], axis=1)
  28. fs = 500
  29. phase = data_str[:, -2].astype(int)
  30. i_trial = data_str[:, -1].astype(int)
  31. edges = np.insert(i_trial[1:i_trial.shape[0]] - i_trial[0:-1], obj=0, values=0)
  32. edges[edges.shape[0] - 1] = 1
  33. hp = signal.butter(2, 1, 'hp', fs=fs, output='sos')
  34. lp = signal.butter(2, 40, 'lp', fs=fs, output='sos')
  35. filtered_data = np.zeros(data.shape)
  36. for i in range(data.shape[1]):
  37. filtered_data[:, i] = data[:, i]
  38. for i in range(data.shape[1]):
  39. filtered_data[:, i] = signal.sosfiltfilt(hp, filtered_data[:, i])
  40. filtered_data[:, i] = signal.sosfiltfilt(lp, filtered_data[:, i])
  41. random_state = 23
  42. ica = FastICA(n_components=None, random_state=random_state)
  43. S_ = ica.fit_transform(filtered_data)
  44. S_plot = np.zeros(S_.shape)
  45. for j in range(S_.shape[1]):
  46. S_plot[:, j] = S_[:, j] + 50 * j
  47. plt.plot(S_plot)
  48. plt.show()
  49. sources_to_delete = [0, 1, 5]
  50. for j in sources_to_delete:
  51. S_[:, int(j)] = 0
  52. filtered_data = ica.inverse_transform(S_)
  53. # bad_trial_margin = 20
  54. # bad_trials2 = np.unique(i_trial[np.where(np.bitwise_or(filtered_data > filtered_data.mean() + bad_trial_margin * filtered_data.std(), filtered_data < filtered_data.mean() - bad_trial_margin * filtered_data.std()))[0]])
  55. threshold = 50
  56. bad_trials2 = np.unique(i_trial[np.where(np.bitwise_or(filtered_data > filtered_data.mean() + threshold, filtered_data < filtered_data.mean() - threshold))[0]])
  57. plt.plot(filtered_data)
  58. plt.plot(phase * np.ptp(filtered_data))
  59. plt.plot(edges * np.ptp(filtered_data))
  60. # plt.plot((filtered_data.mean() + bad_trial_margin * filtered_data.std()) * np.ones(filtered_data.shape[0]))
  61. # plt.plot((filtered_data.mean() - bad_trial_margin * filtered_data.std()) * np.ones(filtered_data.shape[0]))
  62. plt.plot((filtered_data.mean() + threshold) * np.ones(data.shape[0]))
  63. plt.plot((filtered_data.mean() - threshold) * np.ones(data.shape[0]))
  64. plt.show()
  65. plt.show()
  66. i_edges = np.where(edges == 1)[0]
  67. trial_lengths = i_edges[1:i_edges.shape[0]] - i_edges[0:-1]
  68. bad_trial_threshold = 2
  69. bad_trials = np.where(np.bitwise_or(trial_lengths > trial_lengths.mean() + bad_trial_threshold * trial_lengths.std(), trial_lengths < trial_lengths.mean() - bad_trial_threshold * trial_lengths.std()))[0]
  70. trial_length = np.min(np.delete(trial_lengths, bad_trials))
  71. n_trials = i_edges.shape[0] - 1
  72. trial_labels = phase[i_edges[0:i_edges.shape[0]-1] + int(np.round(trial_length/2))]
  73. i_rest = np.where(trial_labels == 0)[0]
  74. i_left = np.where(trial_labels == 1)[0]
  75. i_right = np.where(trial_labels == 2)[0]
  76. print(i_edges)
  77. print(trial_length)
  78. print(bad_trials)
  79. print(bad_trials2)
  80. print(n_trials)
  81. print(trial_labels)
  82. print(i_rest)
  83. print(i_left)
  84. print(i_right)
  85. plt.plot(trial_lengths)
  86. plt.plot((trial_lengths.mean() + bad_trial_threshold * trial_lengths.std()) * np.ones(trial_lengths.shape))
  87. plt.plot((trial_lengths.mean() - bad_trial_threshold * trial_lengths.std()) * np.ones(trial_lengths.shape))
  88. plt.show()
  89. rest_data = np.zeros((i_rest.shape[0], trial_length, data.shape[1]))
  90. for i in range(i_rest.shape[0]):
  91. if not ((i_rest[i] in bad_trials) or (i_rest[i] in bad_trials2)):
  92. rest_data[i, :, :] = filtered_data[i_edges[i_rest[i]]:i_edges[i_rest[i]] + trial_length, :]
  93. rest_data = np.delete(rest_data, obj=np.where(np.sum(np.sum(rest_data, axis=1), axis=1) == 0)[0], axis=0)
  94. left_data = np.zeros((i_left.shape[0], trial_length, data.shape[1]))
  95. for i in range(i_left.shape[0]):
  96. if not ((i_left[i] in bad_trials) or (i_left[i] in bad_trials2)):
  97. left_data[i, :, :] = filtered_data[i_edges[i_left[i]]:i_edges[i_left[i]] + trial_length, :]
  98. left_data = np.delete(left_data, obj=np.where(np.sum(np.sum(left_data, axis=1), axis=1) == 0)[0], axis=0)
  99. right_data = np.zeros((i_right.shape[0], trial_length, data.shape[1]))
  100. for i in range(i_right.shape[0]):
  101. if not ((i_right[i] in bad_trials) or (i_right[i] in bad_trials2)):
  102. right_data[i, :, :] = filtered_data[i_edges[i_right[i]]:i_edges[i_right[i]] + trial_length, :]
  103. right_data = np.delete(right_data, obj=np.where(np.sum(np.sum(right_data, axis=1), axis=1) == 0)[0], axis=0)
  104. # rest_average = rest_data.mean(axis=0)
  105. # left_average = left_data.mean(axis=0)
  106. # right_average = right_data.mean(axis=0)
  107. #
  108. # for i in range(data.shape[1]):
  109. # rest_average[:, i] = rest_average[:, i] - rest_average[0, i]
  110. # left_average[:, i] = left_average[:, i] - left_average[0, i]
  111. # right_average[:, i] = right_average[:, i] - right_average[0, i]
  112. #fig, (ax1, ax2, ax3) = plt.subplots(1, 3)
  113. #ax1.plot(rest_average)
  114. #ax2.plot(left_average)
  115. #ax3.plot(right_average)
  116. #plt.show()
  117. k_lower = int(np.round(8 * trial_length / fs))
  118. #k_upper = int(np.round(int(np.floor(fs/2)) * trial_length / fs))
  119. k_upper = int(np.round(30 * trial_length / fs))
  120. print(k_lower)
  121. print(k_upper)
  122. for i in range(rest_data.shape[0]):
  123. for j in range(rest_data.shape[2]):
  124. rest_data[i, :, j] = np.square(np.abs(fft(rest_data[i, :, j])))
  125. rest_data = rest_data[:, k_lower:k_upper, :]
  126. # for i in range(rest_data.shape[0]):
  127. # rest_data[i, :, :] = rest_data[i, :, :] / np.sum(rest_data[i, :, :])
  128. for i in range(left_data.shape[0]):
  129. for j in range(left_data.shape[2]):
  130. left_data[i, :, j] = np.square(np.abs(fft(left_data[i, :, j])))
  131. left_data = left_data[:, k_lower:k_upper, :]
  132. # for i in range(left_data.shape[0]):
  133. # left_data[i, :, :] = left_data[i, :, :] / np.sum(left_data[i, :, :])
  134. for i in range(right_data.shape[0]):
  135. for j in range(right_data.shape[2]):
  136. right_data[i, :, j] = np.square(np.abs(fft(right_data[i, :, j])))
  137. right_data = right_data[:, k_lower:k_upper, :]
  138. # for i in range(right_data.shape[0]):
  139. # right_data[i, :, :] = right_data[i, :, :] / np.sum(right_data[i, :, :])
  140. rest_data = rest_data[:, :, 2:5]
  141. left_data = left_data[:, :, 2:5]
  142. right_data = right_data[:, :, 2:5]
  143. # for i in range(rest_data.shape[1]):
  144. # for k in range(rest_data.shape[2]):
  145. # avg = (np.average(rest_data, axis=0) + np.average(left_data, axis=0) + np.average(left_data, axis=0)) / 3
  146. # for j in range(rest_data.shape[0]):
  147. # rest_data[j, i, k] = rest_data[j, i, k] / avg[i, k]
  148. # for j in range(left_data.shape[0]):
  149. # left_data[j, i, k] = left_data[j, i, k] / avg[i, k]
  150. # for j in range(right_data.shape[0]):
  151. # right_data[j, i, k] = right_data[j, i, k] / avg[i, k]
  152. # for i in range(rest_data.shape[2]):
  153. # rest_data[:, :, i] = (rest_data[:, :, i] - np.min(rest_data[:, :, i])) / np.std(rest_data[:, :, i])
  154. # left_data[:, :, i] = (left_data[:, :, i] - np.min(left_data[:, :, i])) / np.std(left_data[:, :, i])
  155. # right_data[:, :, i] = (right_data[:, :, i] - np.min(right_data[:, :, i])) / np.std(right_data[:, :, i])
  156. for i in range(rest_data.shape[2]):
  157. rest_data[:, :, i] = (rest_data[:, :, i] - np.min(rest_data[:, :, i])) / (np.max(rest_data[:, :, i]) - np.min(rest_data[:, :, i]))
  158. left_data[:, :, i] = (left_data[:, :, i] - np.min(left_data[:, :, i])) / (np.max(left_data[:, :, i]) - np.min(left_data[:, :, i]))
  159. right_data[:, :, i] = (right_data[:, :, i] - np.min(right_data[:, :, i])) / (np.max(right_data[:, :, i]) - np.min(right_data[:, :, i]))
  160. rest_average = rest_data.mean(axis=0)
  161. left_average = left_data.mean(axis=0)
  162. right_average = right_data.mean(axis=0)
  163. print(rest_data.shape)
  164. print(left_data.shape)
  165. print(right_data.shape)
  166. shuffling_vector_rest = np.arange((rest_data.shape[0]))
  167. shuffling_vector_left = np.arange((left_data.shape[0]))
  168. shuffling_vector_right = np.arange((right_data.shape[0]))
  169. np.random.shuffle(shuffling_vector_rest)
  170. np.random.shuffle(shuffling_vector_left)
  171. np.random.shuffle(shuffling_vector_right)
  172. rest_data = rest_data[shuffling_vector_rest, :, :]
  173. left_data = left_data[shuffling_vector_left, :, :]
  174. right_data = right_data[shuffling_vector_right, :, :]
  175. fig, (ax1, ax2, ax3) = plt.subplots(1, 3, sharey=True)
  176. ax1.plot(rest_average)
  177. ax2.plot(left_average)
  178. ax3.plot(right_average)
  179. plt.show()
  180. training_split = 0.8
  181. training_data = np.concatenate((rest_data[0:int(np.round(training_split * rest_data.shape[0])), :, :],
  182. left_data[0:int(np.round(training_split * left_data.shape[0])), :, :],
  183. right_data[0:int(np.round(training_split * right_data.shape[0])), :, :]), axis=0)
  184. test_data = np.concatenate((rest_data[int(np.round(training_split * rest_data.shape[0])):rest_data.shape[0], :, :],
  185. left_data[int(np.round(training_split * left_data.shape[0])):left_data.shape[0], :, :],
  186. right_data[int(np.round(training_split * right_data.shape[0])):right_data.shape[0], :, :]), axis=0)
  187. avg = np.average(np.concatenate((training_data, test_data), axis=0))
  188. std = np.std(np.concatenate((training_data, test_data), axis=0))
  189. training_data = (training_data - avg) / std
  190. test_data = (test_data - avg) / std
  191. training_data2 = np.reshape(training_data, (training_data.shape[0], training_data.shape[1] * training_data.shape[2]))
  192. test_data2 = np.reshape(test_data, (test_data.shape[0], test_data.shape[1] * test_data.shape[2]))
  193. training_labels = np.zeros((training_data.shape[0]), dtype=int)
  194. training_labels[int(np.round(training_split * rest_data.shape[0])):int(np.round(training_split * rest_data.shape[0])) + int(np.round(training_split * left_data.shape[0]))] = 1
  195. training_labels[int(np.round(training_split * rest_data.shape[0])) + int(np.round(training_split * left_data.shape[0])):training_labels.shape[0]] = 2
  196. test_labels = np.zeros((test_data.shape[0]), dtype=int)
  197. test_labels[int(np.round((1 - training_split) * rest_data.shape[0])):int(np.round((1 - training_split) * rest_data.shape[0])) + int(np.round((1 - training_split) * left_data.shape[0]))] = 1
  198. test_labels[int(np.round((1 - training_split) * rest_data.shape[0])) + int(np.round((1 - training_split) * left_data.shape[0])):test_labels.shape[0]] = 2
  199. clf = svm.NuSVC(nu=0.1, kernel='rbf', class_weight='balanced', decision_function_shape='ovo', tol=1e-3, verbose=True)
  200. clf.fit(training_data2, training_labels)
  201. test_labels_predicted = clf.predict(test_data2)
  202. cm = confusion_matrix(test_labels, test_labels_predicted)
  203. print(cm)
  204. print(accuracy_score(test_labels, test_labels_predicted))
  205. clf = RandomForestClassifier(n_estimators=100, criterion='entropy')
  206. clf.fit(training_data2, training_labels)
  207. test_labels_predicted = clf.predict(test_data2)
  208. cm = confusion_matrix(test_labels, test_labels_predicted)
  209. print(cm)
  210. print(accuracy_score(test_labels, test_labels_predicted))
  211. training_data = training_data.reshape((training_data.shape[0], training_data.shape[1], training_data.shape[2], 1))
  212. test_data = test_data.reshape((test_data.shape[0], test_data.shape[1], test_data.shape[2], 1))
  213. training_labels_CNN = keras.utils.to_categorical(training_labels, 3)
  214. test_labels_CNN = keras.utils.to_categorical(test_labels, 3)
  215. model = keras.Sequential(
  216. [
  217. keras.Input(shape=training_data[0].shape),
  218. layers.Conv2D(8, kernel_size=(3, 3), activation="relu", padding="same"),
  219. layers.Dropout(0.5),
  220. layers.Conv2D(16, kernel_size=(3, 3), activation="relu", padding="same"),
  221. layers.Dropout(0.5),
  222. layers.MaxPooling2D(pool_size=(2, 2)),
  223. layers.Flatten(),
  224. layers.Dropout(0.5),
  225. layers.Dense(3, activation="softmax"),
  226. ]
  227. )
  228. batch_size = training_data.shape[0]
  229. epochs = 2500
  230. model.compile(loss="categorical_crossentropy", optimizer="adam", metrics=["accuracy"])
  231. callback = keras.callbacks.EarlyStopping(monitor='val_accuracy', patience=1000, start_from_epoch=25, restore_best_weights=True)
  232. train_history = model.fit(training_data, training_labels_CNN, batch_size=batch_size, epochs=epochs, callbacks=callback, validation_split=0.25)
  233. test_labels_predicted = np.argmax(model.predict(test_data), axis=1)
  234. cm = confusion_matrix(test_labels, test_labels_predicted)
  235. print(cm)
  236. score = model.evaluate(test_data, test_labels_CNN, verbose=0)
  237. print("Test loss:", score[0])
  238. print("Test accuracy:", score[1])
  239. loss = train_history.history['loss']
  240. accuracy = train_history.history['accuracy']
  241. val_loss = train_history.history['val_loss']
  242. val_accuracy = train_history.history['val_accuracy']
  243. plt.plot(loss)
  244. plt.plot(accuracy)
  245. plt.plot(val_loss)
  246. plt.plot(val_accuracy)
  247. plt.legend(['loss', 'accuracy', 'val_loss', 'val_accuracy'])
  248. plt.show()

MI.py at commit fef74d6, no license · at the source

Overview

Authors: János Rokai1, Melinda Rácz1,2,3, Melinda Becske2,3,4,5, János Csipor6, Csaba Márton Köllőd1,7, István Ulbert1,7,8, Gergely Márton1,6
  1. Institute of Cognitive Neuroscience and Psychology, HUN-REN Research Centre for Natural Sciences, Magyar Tudósok krt. 2, Budapest, 1117 Hungary
  2. School of PhD Studies, Semmelweis University, Üllői út 26, Budapest, 1085 Hungary
  3. Selye János Doctoral College for Advanced Studies, Semmelweis University, Üllői út 22, Budapest, 1085 Hungary
  4. Department of Psychiatry and Psychotherapy, Semmelweis University, Balassa u. 6, Budapest, 1083 Hungary
  5. Nyiro Gyula National Institute of Psychiatry and Addictology, Neurocognitive Research Centre, 59-61 Lehel Utca, Budapest, 1135 Hungary
  6. MindRove Kft., Hédervári út 43, Gyor, 9026 Hungary
  7. Faculty of Information Technology and Bionics, Pázmány Péter Catholic University, Práter Utca 50/a, Budapest, 1083 Hungary
  8. Department of Neurosurgery and Neurointervention, Faculty of Medicine, Semmelweis University, Amerikai út 57, Budapest, 1145 Hungary
Journal: Scientific reports, volume 16, issue 1, article 23147
Dates: received 19 January 2026; accepted 6 May 2026; published online 24 July 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1038/s41598-026-52672-8 · PMID 42493516 · PMCID PMC13396441 · OpenAlex W7170172014
Open access: gold, a free copy (OpenAlex)
Status: code verified (how the code is found and verified)
Categories: EEG (modality), human (organism), methods / tools (subfield)
Methods: Spectral & time-frequency, Statistics, Smoothing, state filtering, decompositions, Machine learning, Preprocessing, Evoked potentials, Physiology & signal measures
Keywords: Electroencephalography, Portable electroencephalography device, Visually evoked potential, Transient visually evoked potential, Event-related potential, P300, Motor execution, Brain–computer interface, Biological techniques, Engineering, Neuroscience
MeSH: Brain-Computer Interfaces*, Electroencephalography*, Convolutional Neural Networks, Electrodes, Event-Related Potentials, P300, Evoked Potentials, Visual, Humans, Signal-To-Noise Ratio, Support Vector Machine (* major topic)
Topic: EEG and Brain-Computer Interfaces (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: HUN-REN Research Centre for Natural Sciences
Citations: cited by 1 paper (Europe PMC); 101 references in the paper

Abstract

In recent years, commercial lightweight electroencephalography (EEG) headsets are gaining popularity in neuroscience. These devices commonly utilize only a few dry electrodes in specific locations and signal quality is often inferior compared to that of their traditional counterparts. In this study, we wanted to assess the feasibility of portable, paste-less, passive electrode-based EEG headset MindRove vision (VSN) for laboratory use. Three paradigms were implemented for acquiring visual evoked potential (VEP), P300 event-related potential and motor execution task (ME) related cortical patterns. Measurements were taken by using VSN, with wet-electrode system mBrainTrain SMARTING applied as reference. The performance of the devices was assessed by using signal-to-noise ratio (SNR) for VEP and P300 while support vector machine, random forest and convolutional neural network-based classifiers were fit to ME data. The SNRdB (i.e. SNR expressed in decibels) of VSN was greater for both VEP and P300, by a margin of 1.998 and 2.845 dB, respectively. There was a significant difference between VEP signal amplitude levels and SNRdB, P300 SNR and SNRdB in favor of VSN. Average accuracy of the sorters were 78.8% for VSN and 80.9% for SMARTING; the difference was not significant. The application of VSN is feasible for use in research besides qualitative exploration.

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.

MelindaRacz/VSN_validation_supplementary

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: fef74d666f998999839cddc889d90e15251c7b12, 6 September 2024
Languages: R (8), Python (3)
Size: 317 files, 11 scripts
Software Heritage: not archived
Found in: “Data availability”
Holds: README, 8 notebooks
Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: Matplotlib (3 files), NumPy (3 files), scikit-learn (3 files), SciPy (3 files), Keras (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
12 files

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;
  • 11 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

Data availability

Measurement data, preprocessing scripts, additional figures and details on statistical analysis are provided as supplementary material to this article, made accessible at (https://github.com/MelindaRacz/VSN_validation_supplementary). Other details are provided upon request addressed to the corresponding author.

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

Versions

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

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 7 authors, 11 keywords, 9 MeSH terms, 1 funder, 72 references.

Cite

This paper

Rokai, J., Rácz, M., Becske, M., Csipor, J., Köllőd, C. M., Ulbert, I., & Márton, G. (2026). Validation of portable, semi-dry electrode-based electroencephalography device for its application in brain-computer interface solutions. Scientific reports, 16(1), 23147. https://doi.org/10.1038/s41598-026-52672-8

BibTeX

@article{rokai2026validation,
author = {Rokai, János and Rácz, Melinda and Becske, Melinda and Csipor, János and Köllőd, Csaba Márton and Ulbert, István and Márton, Gergely},
title = {{Validation of portable, semi-dry electrode-based electroencephalography device for its application in brain-computer interface solutions}},
journal = {Scientific reports},
year = {2026},
month = jul,
volume = {16},
number = {1},
pages = {23147},
publisher = {Nature Publishing Group},
issn = {2045-2322},
doi = {10.1038/s41598-026-52672-8},
url = {https://doi.org/10.1038/s41598-026-52672-8},
pmid = {42493516},
pmcid = {PMC13396441}
}

RIS

TY - JOUR
AU - Rokai, János
AU - Rácz, Melinda
AU - Becske, Melinda
AU - Csipor, János
AU - Köllőd, Csaba Márton
AU - Ulbert, István
AU - Márton, Gergely
TI - Validation of portable, semi-dry electrode-based electroencephalography device for its application in brain-computer interface solutions
T2 - Scientific reports
J2 - Sci Rep
PY - 2026
DA - 2026/07/24
VL - 16
IS - 1
SP - 23147
SN - 2045-2322
PB - Nature Publishing Group
DO - 10.1038/s41598-026-52672-8
UR - https://doi.org/10.1038/s41598-026-52672-8
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s41598-026-52672-8",
"type": "article-journal",
"title": "Validation of portable, semi-dry electrode-based electroencephalography device for its application in brain-computer interface solutions",
"container-title": "Scientific reports",
"author": [
{
"family": "Rokai",
"given": "János"
},
{
"family": "Rácz",
"given": "Melinda"
},
{
"family": "Becske",
"given": "Melinda"
},
{
"family": "Csipor",
"given": "János"
},
{
"family": "Köllőd",
"given": "Csaba Márton"
},
{
"family": "Ulbert",
"given": "István"
},
{
"family": "Márton",
"given": "Gergely"
}
],
"container-title-short": "Sci Rep",
"volume": "16",
"issue": "1",
"page": "23147",
"DOI": "10.1038/s41598-026-52672-8",
"PMID": "42493516",
"PMCID": "PMC13396441",
"ISSN": "2045-2322",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s41598-026-52672-8",
"language": "en",
"issued": {
"date-parts": [
[
2026,
7,
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.pone.0291186
A scoping review on the use of consumer-grade EEG devices for research.
Journal: PloS one
In common: methods / tools, EEG, 10 references
[2] doi:10.7717/peerj.20416 [code]
Evaluating portable EEG: a comparison between two wireless systems (EPOC Flex and LiveAmp) and the wired BrainAmp system.
Journal: PeerJ
In common: EEG, 10 references
[3] doi:10.3758/s13428-025-02905-x [code]
Comparing aperiodic activity in consumer-grade and research-grade EEG: Reliability and association with mathematical ability.
Journal: Behavior research methods
In common: scikit-learn, SciPy, Matplotlib, 1 other tool, EEG, 7 references
[4] doi:10.2196/84707 [code]
Wireless Electroencephalography in Research on Children With Developmental Disabilities: Scoping Review.
Journal: Journal of medical Internet research
In common: EEG, 5 references
[5] doi:10.1016/j.heliyon.2023.e18433 [code]
Crowdsourced EEG experiments: A proof of concept for remote EEG acquisition using EmotivPRO Builder and EmotivLABS.
Journal: Heliyon
In common: SciPy, NumPy, EEG, 4 references
[6] doi:10.1038/s41746-026-03424-5 [code]
Evaluation of PainWaive: a consumer-grade EEG headset for remote EEG monitoring and neurofeedback research in chronic pain.
Journal: NPJ digital medicine
In common: EEG, 5 references
[7] doi:10.3390/s25237262
Traditional Machine Learning Outperforms EEGNet for Consumer-Grade EEG Emotion Recognition: A Comprehensive Evaluation with Cross-Dataset Validation.
Journal: Sensors (Basel, Switzerland)
In common: EEG, 5 references
[8] doi:10.1186/s12859-024-05865-9 [code]
HEROIC: a platform for remote collection of electroencephalographic data using consumer-grade brain wearables.
Journal: BMC bioinformatics
In common: NumPy, methods / tools, EEG, 4 references
[9] doi:10.1371/journal.pone.0354976 [code]
Improved motor imagery BCI performance via task-unaware compression in the BELT Bayesian Edge-Cloud architecture.
Journal: PloS one
In common: Keras, SciPy, Matplotlib, 1 other tool, methods / tools, EEG, 2 references
[10] doi:10.3390/s23187908 [code]
Enhancing Cross-Subject Motor Imagery Classification in EEG-Based Brain-Computer Interfaces by Using Multi-Branch CNN.
Journal: Sensors (Basel, Switzerland)
In common: Keras, scikit-learn, SciPy, 2 other tools, EEG, 2 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.

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/s41598-026-52672-8.svg)](https://openscicode.org/badge/doi/10.1038/s41598-026-52672-8)
<!-- oscr-badge:end -->
reStructuredText
.. oscr-badge:start

.. image:: https://openscicode.org/badge/doi/10.1038/s41598-026-52672-8.svg
   :target: https://openscicode.org/badge/doi/10.1038/s41598-026-52672-8
   :alt: OSCR

.. oscr-badge:end
HTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.1038/s41598-026-52672-8"><img src="https://openscicode.org/badge/doi/10.1038/s41598-026-52672-8.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.