Dissertation > Excellent graduate degree dissertation topics show

Biomedical signal similarity measure research

Author: LianShiLiu
Tutor: ZhengGang
School: Tianjin University of Technology
Course: Applied Computer Technology
Keywords: Biological signal Similarity measure hausdorff Passage
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 105
Quote: 1
Read: Download Dissertation

Abstract


Biological signals are obtained by physical methods, quantitative description of the physical signal human health. It can be an electrical signal, the pressure signal can also be, for example, EEG, ECG, pulse signals, doctors usually based on these changes in a biological signal diagnose various diseases. With the development of computer technology, these signals through a discretized into a vector by a variety of intelligent processing technology analysis, assisted medical diagnostics. The key elements of the biological signal analysis is to describe the degree of similarity between the signals, which directly affects the accuracy of the measure medical diagnosis. However, given the biological signal is accompanied by noise, baseline drift, the amplitude of contraction, bending, etc. interfere with the timeline, as well as their reasons for weak signals, the traditional similarity measurement methods used in biological signal similarity measure calculated effect is not very satisfied. In this paper, in-depth understanding of the characteristics of biomedical signal from the signal timing and curve shape to start, we propose a biological signal characteristics for similarity measure between signals and research from the following three aspects: This article is based on biological signals Medicine and timing characteristics and scientific analysis of the key information representative of a biological signal feature points, and based on these characteristic points of a biological signal segments, then each signal contribution of the diagnosis given different weights, the similarity measure for the integration in the provide pre-prepared formula. Based on the curve shape of a biological signal of the research, the binary image will be used to measure the similarity of the Hausdorff distance is used to measure the similarity of biological signals, effectively overcome due to noise, baseline drift, the amplitude of contraction, bending axis caused by the change of the signal curve. Meanwhile, in order to improve discrimination Hausdorff distance, the Hausdorff distance calculation method is improved, about the feature points into a biological signal, such as pre-assigned weights for calculating the Hausdorff distance, so that the results can be reflected more biosignal curve characteristics. Through the MIT / BIH database experimental data show that the method has better discrimination. In order to calculate the similarity measure is further reflected biological signals characteristic curve, this paper Hausdorff distance based on the idea proposed similarity measure based channel strategy. It is through a graphical overlay method known class data sets, making the signal curve similar to the formation of the icon of the channel, and then use image processing method to determine the channel boundaries, so as to achieve determine unknown signal-to-channel attribution. Experimental results show that this method can effectively distinguish between different forms of biological signals, and a certain degree of tolerance. In this paper, the content and the results obtained with a wide range of application value. It not only can be used to measure the similarity between a biological signal, a biological signal can be used to construct class template library, the evaluation of biological signals clustering results, based on a biological signal identification and other fields. In addition, the article also has a non-linear curve data timing characteristics, the icon of the classification methods proposed ideas and specific tests, for the further study of biological signals provides a new way of thinking.

Related Dissertations

  1. Research on Vision Detection Algorithm for Tracing Printing System,TP391.41
  2. Finger vein recognition technology,TP391.41
  3. The Research of Image Matching Technology Based on Feature Detection,TP391.41
  4. Image Matching Based on Local Invariant Features,TP391.41
  5. Holter waveforms clustering strategy effectiveness analysis,TP311.13
  6. Technologies of Real-time Traffic Sign Segmentation and Recognition,TP391.41
  7. Polarimetric SAR coherent information on registration and change detection algorithm,TN957.52
  8. Vision Position and Control of Plane Work Piece Based on Adept Robot,TP391.41
  9. Traffic statistics based on moving target tracking technology,TP391.41
  10. Moving Objects Detection and Tracking of Video Images in Industrial Sewing Machines,TP391.41
  11. Research on Palm Vein Recognition Algorithm,TP391.41
  12. The Method of Auto-annotation for the Metairie of Cartoon Character,TP391.41
  13. On Pixel-level Image Registration and Fusion Based on Wavelet,TP391.41
  14. Research on Parallel Algorithms of Image Matching,TP301.6
  15. Study on Close-Range Stereo Image Matching Method Based on Nonsampled Contourlet Transform,TP391.41
  16. Effects of Visfatin on Ion Channels in Rat Cardiac Ventricular Myocytes,R54
  17. Opening Double-door World,I207.42
  18. Harbor Target Detection and Recognition in Remote Sensing Image,TP751
  19. The Research of Vehicle Recognition Based on Video Sequences,TP391.41
  20. General distribution interval symbolic data clustering analysis method,TP311.13
  21. Near Infrared Hand Vein Recognition Algorithm,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile