Electromyographic data recordings during isometric contractions of the biceps brachii and deltoid collected from patients and healthy subjects.
Description
Electromyography (EMG) is a valuable tool for exploring both the nervous and muscular system. Two forms of pathologies called myopathy and neuropathy can affect the neuromuscular system. Myopathy refers to muscle conditions where muscle fibers show dysfunction while neuropathy is a neurological disorder related to the degeneration of nerves, resulting in muscle weakness, cramps, stiffness, muscle wasting, etc. Therefore, needle electromyography, known as intramuscular EMG (iEMG), is mainly used in clinical practice to diagnose patients with neuromuscular disorders. The dataset presented in this article contains three classes of raw invasive EMG signals, recorded from two nature of muscles, biceps brachii and deltoid, this data was collected from 241 participants, in hospital environment, spanning various ages and genders, 191 patients afflicted with either neuropathy or myopathy disease and 50 healthy subjects during isometric contraction. Furthermore, a Matlab code is provided to facilitate the reading of EMG files, carry out resampling and filtering procedures, and eliminate the 50 Hz noise component. This dataset aimed to compare healthy and non-healthy subjects. Therefore, it can be useful as a reference to build an artificial intelligence model or validate a machine learning models created with different datasets, for the early detection of neuromuscular pathologies. Experimental dataset is considered of raw invasive EMG signals, this dataset were collected from October 2015 to April 2019 at the Neurology Department of Mustapha Pacha University Hospital of Algiers. These signals were collected with collaboration and support of specialist clinicians throughout the process and in partnership with the Spoken Communication and Signal Processing Laboratory and Instrumentation Laboratory of the Faculty of Electronics and Informatics of Houari Boumediene University of Sciences and Technology. This EMG dataset covers a large population of healthy people in very good shape and sick subjects of various ages, genders, and neuromuscular pathologies like myopathy (i.e., myotonia and polymyositis) and neuropathy (i.e., anterior horn and multiple sclerosis and radicular damage), showing different severity level of disease
Files
Steps to reproduce
Data was collected from three distinct classes of subjects, the normal control group consist of 50 healthy volunteer subjects, of which 23 are females and 27 are males, aged between 17-56 years old (mean of 27.9 ± 9.5), chosen randomly, and all of them are in general good physical shape and none had signs or history of neuromuscular or musculoskeletal disorders. The second group is composed of 98 myopathy subjects, 55 females and 43 males, aged between 4-78 years old (mean of 41.6 ± 16.8). In the neuropathy group, there are 93 subjects, 53 females and 40 males, aged between 2-81 years old (mean of 46.5 ± 16.3), The data acquisition system includes the MATRIX LIGHT 2 channel EMG EP Headbox with Integrated Electric Simulator manufactured by Micromed SPA, disposable concentric needles electrodes with golden connectors from MYOLINE brand, and SystemPlus Evolution 1.04 software. Acquired EMG signals have amplitude in millivolts within a frequency band of [16Hz, 5000Hz]. The sampling frequency is 32768Hz. The dataset directory structure consists of two main folders, one containing data related to the deltoid muscle and the other containing data for the biceps brachii muscle. Each folder is further divided into three subdirectories, corresponding to healthy subjects, myopathy patients and neuropathy patients respectively. Within these subdirectories, EMG signal recordings are stored as .asc files. The naming scheme for these files adheres to a specific pattern: they begin with "EMG", followed by the assigned EMG recording number, then the corresponding patient's identification number, followed by the specific muscle group involved (such as RB: right biceps, LB: left biceps, RD: right deltoid, LD: left deltoid), and finally, the class assignment (Healthy: Hea, Myopathy: Myo, Neuropathy: Neu). For instance, the designation "EMG004_02_LD_Hea" signifies the fourth EMG recording corresponding to the patient number 2, recorded from the left deltoid muscle, and belonging to the Healthy class (class 1).