Description

To record activity from skeletal muscle, the ElectroMyoGram (EMG) is used. This technique measures the effect of action potentials spreading over the muscle. In the case of surface EMG, the sum of action potentials from a large part of the muscle is recorded, whereas you can do localized measurements using a needle electrode, i.e. intramuscular EMG or needle EMG. In this task you will analyze three somewhat idealized, noise-free, intramuscular EMG signals. All signals are recorded with a sampling frequency of 25 kHz and contain motor unit action potentials (MUAP) from one or more motor units. Parts of these signals are shown in the figure on the next page. They are found in the file EMG.mat. Your task is to develop an algorithm that identifies the firing pattern in these three signals, i.e. to determine when the MUAP:s arise in time and to distinguish between MUAP:s originating from different motor units. From the firing pattern, calculate the firing rate of each motor unit and track how this firing rate changes over time. Also present the shape of the MUAP from each motor unit. You are recommended to read chapter 5 in Sörnmo and Laguna for solving this task, especially sections 5.1 and 5.6. Sections 5.6.1 and 5.6.2 are advanced and you may very well manage this task fairly good without fully understanding those sections. To help you with your algorithm, the following steps are a good start: 1. Identify the location of each MUAP using some kind of peak detector. 2. Go through the whole signal again and compare the morphology of each MUAP to the MUAP:s detected and stored earlier in the signal. If the examined MUAP is not similar to a previously stored one, store that MUAP as a new. 3. To calculate the firing rate of each motor unit, calculate the time gap between MUAP:s originating from the same motor unit. To give you a further push into this task, we can reveal that the first EMG signal, EMG1, consists of MUAP:s from only one motor unit, which has a constant firing rate during the whole signal. The other two signals, EMG2 and EMG3, are more complex in their nature. To pass this task, your algorithm should be able to solve at least EMG1 and EMG2 in a proper way, whereas a complete resolution of all MUAP:s in EMG3 is difficult. However, even if your algorithm does not entirely succeed with EMG3, you should discuss the difficulties in this signal and in other intramuscular EMG:s in a proper way. You should also discuss the possible physiological background of EMG2 and EMG3, based on the firing rate and motor unit recruitmen

Instruction Files