NMT-5k

Citation Author(s):
Hira
Masood
National University of Sciences and Technology (NUST), Islamabad, Pakistan
Muhammad Imran
Malik
Muhammad
Athar
Muhammad Arsalan
Khan
Saima
Shafait
Faisal
Shafait
Hassan Aqeel
Khan
Submitted by:
Hira Masood
Last updated:
Tue, 02/18/2025 - 10:47
DOI:
10.21227/7qxh-q986
Data Format:
.edf
License:
Creative Commons Attribution
36 Views
Categories:
Machine Learning
Neuroscience
Brain
Keywords:
Deep Learning, Electroencephalography (EEG)
0
0 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

EEG signals in the NMT-5k dataset were captured using a 21-channel setup following the 10/20 electrode placement system, maintaining a sampling rate of 200 Hz. Data collection was conducted using the KT88-2400 system from Contec Medical Systems.

The recordings were annotated by neurological experts, who labeled each recording as normal or pathological based on clinical diagnoses. To ensure high accuracy and consistency, the labels were verified through a review process involving two neurologists, and only records with consensus labels were included in the dataset. The dataset employs the same linked ear reference montage as the original NMT dataset.

Instructions: 

The dataset contains EEG signal data in EDF file format, comprising recordings from 21 channels. The data is organized into two folders: one for normal cases and another for abnormal cases. Each file is named according to the patient's record number.

Dataset Files

    Files have not been uploaded for this dataset

     

    QUESTIONS?