Study of various Data Compression Techniques used in Lossless Compression of EEG Signals
- 1. STUDY OF VARIOUS DATA COMPRESSION TECHNIQUES USED IN LOSSLESS COMPRESSION OF EEG SIGNALS Author: ANKITA TIWARI Amity School of Engineering and Technology Amity University, Uttar Pradesh
- 2. INTRODUCTION • As we know that developments in technology are introducing various methods for Tele-medicine. • Tele-medicine includes many of the applications and this is one of the fields in telemedicine which have seen excellent growth. • In the procedures of Tele-medicine we record a extremely large amount of EEG real time data. • Therefore we require an efficient and lossless technique that is able to perform compression of recorded EEG signals.
- 3. • In this paper we have studied and analysed various lossless data compression techniques used in the compression of EEG signals. • In the course of studying various techniques we have presented the analysis of some most widely used time domain techniques those are AZTEC (Amplitude zone time epoch coding) technique and Turning point technique (TP) and in transformation based compression techniques we have presented the study of Discrete Cosine Transform technique (DCT) performed with Huffman coding technique and Empirical Mode Decomposition (EMD) technique
- 4. • The overall performance of all these techniques are studied and analysed on the basis of two main parameters those are the compression ratio(CR) and Percent Root means square Difference(PRD). • We have used the data base of BCI website for the calculation of CR and PRD
- 5. Amplitude Zone Time Epoch Coding (AZTECH) • J.R. Cox et al have introduced a technique named as the AZTEC • In this technique ECG signal is converted in to the slopes and various horizontal lines called Plateaus • In these Plateaus we uses a interpolator of zero- order.in this compression technique the information contained by the signals is stored in the form of amplitude of the samples and the length obtained through captured signals
- 6. • A Plateau forms a slope in the condition when the length is less than three. • In AZTEC we obtains the compression ratio up to 10:1 but it is not a well suitable technique for the ECG data compression because in the decompression the signal is reconstructed in the form of steps that is not useful way for the diagnosis
- 7. Turning point (TP) Technique • In this technique first step we performs the recording of EEG signals • After the recording of the signal we take out 3 points from the captured signal and performs the below checking method: (T1-T0)*(T2-T1)<0 Or (T1-T0)*(T2-T1)>0 • We will store point T1 if from above conditions, condition one comes true otherwise we will store point T2 • Decompression of the compressed ECG signals is performed
- 8. Discrete Cosine Transform (DCT) with Huffman coding technique • DCT technique which is going to be used for the compression is transform type technique of data compression • that this technique is quite similar to the DFT • In this technique the transformation of the recoded EEG signal is done into frequency representation from spatial representation. • During our study we have analysed that this technique compresses and represents the signal in the summation of the changing magnitude and frequency of the signal.
- 9. Empirical Mode Decomposition (EMD) technique • Haung et al has developed a data compression technique named as EMD • This is a transform method of compression • In this technique we first transform our signal into frequency domain and then we perform the compression • In this technique we breakdown our complicated EEG signal using a technique called adaptive decomposition . These decomposed finite no of parts of the original signal are known as the IMF (intrinsic mode functions
- 10. • Using an Hilbert transform based application on these IMF, we obtains the better frequency estimation because in the IMFs we have the zero-mean frequency and amplitude modulated portions • In reconstruction of signal all the IMFs are going to be superimposed over each other along with residual slow that results in the better reconstruction of signals
- 11. FACTORS USED FOR THE ANALYSIS OF COMPRESSION 1. CR = 𝑵 𝒊𝒏 𝑵 𝒐𝒖𝒕 2.
- 12. RESLUTS • The following table compares the cr and prd value for different techniques Technique Parameters TP AZTEC DCT and Huffma n Coding EMD CR 2:1 1.0403 :1 12.750 9:1 22.85: 1 PRD 0.156 7 9.4903 3.9033 2.0730
- 13. Advantage of compression • Reduces the Requirement of the memory space • Compression results in less data for the transmission and reception that reduces the power requirements in the tele medicine • Reduces the cost for the analysis • Analysis becomes easier
- 14. Thank you !