Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống
1
/ 28 trang
THÔNG TIN TÀI LIỆU
Nội dung
Hanoi University of Science and Technology Telecommunication and Electronics Department PERSONAL A U T H E N T I C AT I O N BY SINGLECHANNEL ECG 4/28/2023 STUDENTS: CLASS: VU THI MINH BME K58 INSTRUCTOR: DR NGUYEN VIET DUNG PURPOSES ✓Research about biometric using ECG signal ✓To authenticate a person lead to identify person in the future 4/28/2023 OUTLINE • Background information • Block diagram • ECG acquisition • Pre-processing • Feature extraction • Classification • Results • Conclusion 4/28/2023 BACKGROUND INFORMATION Biometric authentication: "Are you indeed Mr or Mrs A?" Biometric identification "Who are you?" ❑False Acceptance Rate (FAR) FAR = Number of False Acceptances Number of Identification Attemps ❑False Reject Rate (FRR) FRR = 4/28/2023 Number of False Rejections Number of Identification Attemps BACKGROUND INFORMATION Bio- signal A material carrier of the information about the state of the analyzed biological systems Give more detailed characteristics about the system ECG Non- electric biosignals 4/28/2023 Electrical bio-signals BACKGROUND INFORMATION ❑ELECTROCARDIOGRAM (ECG) ▪ f: 0.05 Hz- 100 Hz ▪ A: 1- 10 mV ( dynamic range) ▪ peaks and valleys: P, Q, R, S, T 4/28/2023 BACKGROUND INFORMATION ❑ELECTROCARDIOGRAM (ECG) ▪ PR: 0.12- 0.25s ▪ QRS: 0.08- 0.12s ▪ QT: 0.35-0.44s ▪ ST: 0.05- 0.15s 4/28/2023 BLOCK DIAGRAM Classification Feature extraction Pre- processing Find features ECG Record Classify ‘A’ or ‘not A’ P, Q, R, S,T peaks Record from Kardia mobile Filter to remove noises Statistical data Digitize data 4/28/2023 ECG ACQUISITION ❖ Kardia mobile o To converts electrical impulses from fingertips into ultrasound signals transmitted to the mobile device’s microphone o Specifications 4/28/2023 ECG Channels Single Channel Input Dynamic Range 10 mV Frequency Response 0.5Hz - 40 Hz A/D Sampling Rate 300 Hz Resolution 16 bit Heart Rate Range 30- 300 bpm Battery Type 3V Coin Cell Battery life 12 months typical use ECG ACQUISITION ❖ Kardia mobile 4/28/2023 10 PRE-PROCESSING ❑Filter: Band-pass filter • High-pass filter: 0.05 and 0.5 Hz (low-frequency cutoff ) • Low-pass filters : 40, 100, and 150 Hz (high-frequency cutoff) Filter configuration 0.05–40 Hz 0.5–40 Hz 0.05–100 Hz 0.5–100 Hz 0.05–150 Hz 0.5–150 Hz Choose ❑Cutting segments: from 20s to 50s 4/28/2023 14 PRE- PROCESSING ❑ Pre-processing : Chebychev band-pass filter 0.5- 40 Hz The power spectrum of original signal 4/28/2023 The power spectrum of filtered signal 15 FEATURE EXTRACTION Scaling function of Daubechies Wavelet: A progression {αk; kϵZ} satisfying the following four conditions for all integer N≥2: The expression relating the mother wavelet to the scaling function is: 4/28/2023 16 FEATURE EXTRACTION Daubechies 4/28/2023 Daubechies 17 FEATURE EXTRACTION Decomposition level • S: original signal/ time series data • Ai: approximation low frequency content • Di: Detail high frequency content • Level decomposition: S = A1+ D1 • Level decomposition 2: S = A2 + D2 +D1 • Level decomposition n- level: S = An + Dn +Dn-1 + Dn-2+…+ D1 4/28/2023 18 FEATURE EXTRACTION • Find R peak in decomposition level 4/28/2023 Find peak >= 60% max value Find peaks inversely on original signal From R peaks, find the other peaks based on duration of them Let first peak be R peak Find the other peaks based on the minimum and maximum Mean values of P, Q, R, S, T peaks 19 FEATURE EXTRACTION o Mean o Skewness: a measure for the degree of symmetry in the variable distribution o Median absolute deviation (MAD) o Standard deviation (SD) 4/28/2023 o Kurtosis: a measure for the degree of tailedness in the variable distribution 20 CLASSIFICATION Classification Unsupervised classification Supervised classification Support vector machine (SVM) 4/28/2023 K- Nearest Neighbor 21 RESULTS o Total data: 150 samples / 60 samples of authenticated person o Feature extraction: 11 features ✓ 120 training data : 60 data of authenticated person, data/ each other ( total 60) ✓ Mean, Median, SD, MAD, Skewness, Kurtosis ✓ MeanP, MeanQ, MeanR, MeanS, MeanT ✓ 32 testing data: the ratio 16 /16data 4/28/2023 22 RESULTS Accuracy of train and different validate Classification Train Validate (Test/ train ratio) 10/90 20/80 30/70 40/60 100% 90.9% 91.3% 94.3% 93.5% 99.1% 90.9% 91.3% 97.1% 91.3% KNN SVM Test/ train ration: 30/ 70 Weighted KNN, Medium Gausisan SVM 4/28/2023 23 RESULTS Trained model: Medium Gaussian SVM 4/28/2023 Trained model: Weighted KNN 24 RESULTS Results of test 30 samples with 15 samples of authenticated person Trial SVM KNN 87.50 % 78.13% 0.125 0.125 0.000 0.094 Accuracy (%) FAR FRR 4/28/2023 25 CONCLUSION ❑Classification accuracy of SVM is higher than one of KNN ❑Base on biometric criteria: KNN is better ❑Some causes affect to results: ▪ Small number of samples ▪ Reconstruction ▪ Timer for cutting segments ▪ Feature extraction method 4/28/2023 26 THANK YOU 4/28/2023 27 REFERENCES o Joseph D B.( n.d ) Biomedical engineering fundamental (3rd edition) Trinity college, Hartfort, Conecticuit ,U.S.A o https://www.gemalto.com/govt/inspired/biometrics o http://cimss.ssec.wisc.edu/wxwise/class/aos340/spr00/whatismatlab.htm o Ruben J.M.D , Neson E.V.P, Erika U.,Wavelet Daubechies (db4) Transform Assessment for WorldView-2 Images Fusion, Distrital University Francisco José de Caldas, Carrera No 40B – 53, Bogotá D.C., Colombia o https://heimdalsecurity.com/blog/biometric-authentication/ o https://www.mathworks.com/discovery/pattern-recognition.html o Hargrove, L.; Zhou, P.; Englehart, K.; Kuiken, T.A The effect of ECG interference on pattern-recognition-based myoelectric control for targeted muscle reinnervated patients IEEE Trans Bio Med Eng 2009, 56, 2197–2201 o http://matlab.izmiran.ru/help/toolbox/wavelet/ch01_i15.html o Zardoshti-Kermani, M.; Wheeler, B.C.; Badie, K.; Hashemi, R.M EMG feature evaluation for movement control of upper extremity prostheses IEEE Trans Rehabil Eng 1995, 3, 324–333 o Englehart, K.; Hudgins, B.; Parker, P.A.; Stevenson, M Classification of the myoelectric signal using time-frequency based representations Med Eng Phys 1999, 21, 431–438 o Phinyomark, A.; Phukpattaranont, P.; Limsakul, C Feature reduction and selection for EMG signal classification Expert Syst Appl 2012, 39, 7420–7431 4/28/2023 28