Imran Goker

Classification of Juvenile Myoclonic Epilepsy Data Acquired Through Scanning Electromyography with Machine Learning Algorithms

In this paper, classification of Juvenile Myoclonic Epilepsy (JME) patients and healthy volunteers included into Normal Control (NC) groups was established using Feed-Forward Neural Networks (NN), Support Vector Machines (SVM), Decision Trees (DT), and Naïve Bayes (NB) methods by utilizing the data obtained through the scanning EMG method used in a clinical study. An experimental setup was built for this purpose. 105 motor units were measured. 44 of them belonged to JME group consisting of 9 patients and 61 of them belonged to NC group comprising ten healthy volunteers. k-fold cross validation was applied to train and test the models. ROC curves were drawn for k values of 4, 6, 8 and 10. 100% of detection sensitivity was obtained for DT, NN, and NB classification methods. The lowest FP number, which was obtained by NN, was 5.

Large motor unit territories by scanning electromyography in patients with juvenile myoclonic epilepsy.

Juvenile myoclonic epilepsy is a genetically inherited disorder characterized by myoclonic jerks and generalized seizures. It has been proposed that patients with juvenile myoclonic epilepsy have larger motor units (MUs) than normals by MU number estimation and macro electromyography techniques. In this study, an experimental setup for scanning electromyography was built to investigate electrophysiologic cross-sections of the MU territories in 9 patients with juvenile myoclonic epilepsy, 3 patients with spinal muscular atrophy, and 10 healthy volunteers. Scanning electromyography was performed on the biceps brachii muscle. For each MU, three-dimensional maps of the MU territories were plotted. The length of MU cross-section and the maximum amplitude of each MU were measured from these maps and compared among the three groups of subjects. Like spinal muscular atrophy patients, patients with juvenile myoclonic epilepsy had significantly larger MU territories than normal controls.

Feature extraction and classification of neuromuscular diseases using scanning EMG

In this study a new dataset are prepared for neuromuscular diseases using scanning EMG method and four new features are extracted. These features are maximum amplitude, phase duration at the maximum amplitude, maximum amplitude times phase duration, and number of peaks. By using statistical values such as mean and variance, number of features has increased up to eight. This dataset was classified by using multi layer perceptron (MLP), support vector machines (SVM), k-nearest neighbours algorithm (k-NN), and radial basis function networks (RBF). The best accuracy is obtained as 97.78% with SVM algorithm and 3-NN algorithm.

Interface design for automation of the scanning EMG method

Scanning EMG is a method developed for examining the electro-physiological cross section and the size of the motor unit of a human muscle. Electrical specifications of the motor unit can be obtained as well as anatomical distribution of muscle fibers and pathological changes between different muscles can be examined by the help of this method. In this paper; an automation system which is designed for the execution of scanning EMG method, whether manually or automatically, and a user-interface are described. Parameters like step count and step size which are about the movement of an electrode, moved by a linear actuator, through muscle fibers can be defined as reference by user via designed interface. As a result, acquired signals are digitalized by data-acquisition card (DAQ) and saved as text file for the future signal process tasks.

Determining MUAP activity corridor in scanning EMG recordings

In scanning EMG recordings there are more than one motor unit activity which is recorded by concentric needle electrode. Especially for determining features of a motor unit action potential (MUAP) such as phase duration and number of peaks easily, first the other motor unit activities must be discarded. In this study a new method is proposed to determine the activity corridor that related with the motor unit to be examined in scanning EMG recordings. This method is comprised of wavelet transform based noise reduction and autocorrelation function based location detection. Number of 34 scanning EMG recordings was tested by using this method and the activity corridors were determined correctly.

Profesyonel futbolcularda egzersiz öncesi ve sonrası iskemi modifiye albumin düzeyleri

Objectives: The ischemia-modified albumin (IMA) test is a promising indicator of myocardial ischemia in the early diagnosis of the acute coronary syndrome (ACS). Recent studies examining alterations in IMA levels caused by myocardial ischemia that develops during exercise have shown contradictory results. There has been an increase in severe exercise-related cardiac events in football. We therefore, assessed IMA levels before and after exercise in professional soccer players in order to examine the effect of the exercise on IMA levels. Methods: Blood was collected from professional soccer players before and after exercise. IMA levels were measured by using spectrophotometric methods, and the albumin levels were measured in an autoanalyzer. Results: The average pre- and post exercise IMA levels of 16 soccer players were found to be 0.438±0.071 and 0.386±0.069, respectively. The pre- and post exercise albumin levels of the same group were 4.08±0.20 g/dL and 4.21±0.19 g/dL, respectively. There was a strong negative correlation between post exercise IMA and albumin levels, which was statistically significant (r = -0.535. p = 0.033).Conclusion: IMA levels should be examined not only in noncardiac pathologies, but also in specific serum albumin concentrations and in individuals practicing vigorous sport activities

Motor unit territories in juvenile myoclonic epilepsy patients

In this study, the scanning EMG technique was implemented to investigate electrophysiological cross-sections of the motor unit (MU) territories in healthy volunteers and in subjects with juvenile myoclonic epilepsy and spinal muscular atrophy. Measurements were taken intramuscularly by means of two concentric needle electrodes from biceps brachialis muscles. 3-D maps of the MU territories were plotted for each MU to determine the lengths of MU cross-section and the maximum amplitudes of each MU. There was evidence of a preponderance of large MUs in patients with juvenile myoclonic epilepsy.

S101. A novel automatized F-wave MUNE method: A preliminary analysis

Introduction The aim of this study was to test a novel automatized F-wave MUNE method by comparing its results with those elicited with the manual method performed in patients with amyotrophic lateral sclerosis (ALS) and healthy controls. Methods Ninety F-waves elicited with supramaximal stimuli were recorded from thenar and hypothenar muscles in 10 ALS patients and 4 healthy controls. For manual analysis, F-waves with the same amplitude, latency and shape were selected visually by the same examiner. For automatized analysis, similar signals were grouped according to latencies of negative and positive peaks and peak-to-peak amplitudes in MATLAB by using the original software developed for this study. The mean amplitudes of F-waves repeating more than once were defined as sMUPs and MUNE value for each case was calculated by dividing the CMAP amplitude by mean sMUP amplitude. Results Data of two healthy controls and 1 ALS patient were excluded from the analysis since no repeater F-waves were detected by one of the methods. sMUP and F-wave MUNE values calculated by both methods were found to be highly correlated (Spearman’s rho, p μ V and 406.55  μ V, respectively in thenar muscles and 412.53  μ V and 332.77  μ V in hypothenar muscles). Similar F-wave MUNE values were also calculated by using both methods (26.00 and 28.58, respectively in thenar muscles and 36.04 and 30.18 in hypothenar muscles). Conclusion Our automatized F-wave MUNE method seems to be promising in reflecting motor unit numbers in the hand muscles in patients with ALS and healthy controls.

T78. CMAP scan and scanning EMG in the same muscle: Two cases with post-polio muscular atrophy

Introduction Post-polio muscular atrophy (PPMA) is characterized by new onset or increased weakness in patients with prior poliomyelitis after a stable period of time. Loss of highly reinnervated motor units during ageing has been accused for the development of this syndrome which is also known as “unstable-polio”. These patients have less number of motor units that can be estimated by conventional electrophysiological methods. By showing the large steps, CMAP scan provides information about the amount of collateral reinnervation in the construction of total muscle response (CMAP). As well as the number of motor units, their territory is also an object of curiosity. It is possible to record bioelectrical activity of motor unit lengthwise by scanning EMG and depict the temporal and spatial features of motor unit action potential (MUAP). This presentation aims to combine the findings in CMAP scan with scanning EMG and draw attention to reinnervation status of 2 PPMA patients whose tibialis anterior (TA) muscles were affected in different degrees. Methods Two patients aged 39and 41 years were included. Patient 1 had PPMA for 8 years and his TA muscle strength was 3-/5, whereas Patient 2 showed PPMA findings for 1 year and his TA strength was 4/5. CMAP scan of TA muscle on recently affected side was performed with a commercially available software. In scanning EMG, MU territories were scanned with a concentric needle electrode (CNE) which is attached to a stepper motor. Another CNE is used for sweep triggering with the rate of selected motor units’ firing frequency. Acquired signals were processed by the dedicated software designed by the authors. Results CMAP scan of Patient 1 revealed a 1.65 mV CMAP constituting of 4 very large steps and Patient 2 revealed a 5.5 mV CMAP containing a few smaller steps. In scanning EMG, both patients’ motor units showed increased voltage in different parts corresponding to dense areas arisen from collateral reinnervation. Interestingly, the patient with more pronounced weakness for a longer period revealed both huge steps in his CMAP scan and also showed electrically silent areas in his scanned motor units. On the other hand, the patient with stronger TA muscle did not show very large steps or silent areas in his CMAP scan and MU scan, respectively. Conclusion Loss of dense motor units leads to PPMA. However, in PPMA patients with severe weakness which is depicted by less number of motor units and presence of huge steps in CMAP scan, loss of fractions in motor unit territory might be a principal contributing factor which can only be demonstrated by scanning EMG.

Classification of neuromuscular junction and tendon recordings of neuromuscular diseases by their spectrogram

In this study, the effect of spectrograms from neuromuscular junction and tendon records for normal, neurogenic and myopathic motor units being constructed via EMG Simulator v3.6 on the differential diagnosis were investigated. Multi-layer perceptron is chosen as classifier. If only the neuromuscular junction records are applied to the network, the performance is 73.33%. If only tendon records are applied to the input of network, the performance is 94.67%. When neuromuscular junction and tendon records are applied together to the network, the performance is 100%.