In recent years,immune-based intrusion detection has become a key research in computer security area,which has opened up a new conjecture for computer security.Genetic algorithms,which simulates the natural evolution,can optimize the searching process.Based on natural immune theory,this research simulates natural cognition and memory mechanism to form and analyze an algorithm of constructing detector.

The classical deformable models and some new approaches for the past few years are surveyed,especially de-scribe two important methods of them: snake model and level-set model.Image segmentation is the bases of 3D recon-struction technology and medical visualization image,which are meaningful to disease diagnosis and therapy in clinical.It is not only a key step and critical technology in medical image processing and image analysis but also a classic puzzle.With the need of application,it is very important to continually research the image segmentation,to increasingly improve the old approaches and introduce the new and more effective ones.

BACKGROUND:In spectral CT imaging study, the selection of scanning parameters is considered by most researchers, but the effects of measuring position are often overlooked. Actual measurement found that different measurement location had significant impact on the result of the measurement. Through measurement and mathematical model of a large amount of data, we can correct the measurement results of different location. The results with real data alignment are higher. OBJECTIVE:To explore effects of different measuring positions on results in spectral CT imaging, and to optimize correction coefficients. METHODS:GE standard water phantom was applied to orderly obtain 5 combined scanning parameters with 552 groups of data. Size measurement method was divided into two measuring range:ROI1 and ROI2. We selected 10 points to measure CT value, including Center, North, South, West and East. The measurement data of ROI1 and ROI2 were classified and screened based on the same sequence. Mathematical modeling and probability statistics analysis were used to optimize correction coefficient, get calibration function and draw experimental simulation curve. RESULTS AND CONCLUSION:(1) Measuring methods of ROI1 were superior to the ROI2’s on water phantom in spectrum CT. (2) To different scanning sequences, the measuring results were different. To the same scanning sequences, the measurements for different positions on water phantom in spectrum CT had remarkable influence on the measuring results, which varied from points to points. (3) Through setting up mathematical modeling, using method of statistical analysis, we could get the correction function on different measuring positions. (4) Above results confirmed that compared with the theoretical model and the experimental data of spectral CT scanning parameters, the coefficients of position can be adjusted, which can optimize the measuring results.