Search

In the present paper, a general multivariate approach, estimating cancer latency using data from a case-control study with time-related factors, is proposed based on the theory for risk analysis of states. Making use of a forward-analysis and backward-synthesis strategy and excess exposure fraction, this approach can simultaneously accomplish multivariate logistic regression analysis and estimation of susceptible exposure ages and latency distribution. The latency is resolved into a accumulative period and preneoplastic period. It gives a theoretical interpretation for primary and secondary prevention of cancer.

In this paper, we propose an iterative algorthm called FAHE., which is based on the relativity between the current local histogram and the one before the sliding window moving. Comparing with the basic AHE, the computing time of FAHE is decreased from 5 hours to 4 minutes on a 486dx/33 compatible computer, when using a 65 x 65 sliding window for a 512 x 512 with 8 bits gray-level range.

Image space reconstruction algorithm (ISRA) is a new kind of method for ECT reconstruction. Unlike the ML-EM algorithm which maximizes likelihood function of Possion distribution, ISRA searches for the minimum non-negative mean square solution. In this paper, an improved ISRA (IISRA) algorithm is proposed. A comparison between the algorithms (ISRA and IISRA) demonstrates that the result of IISRA is closer to the test image than that of ISRA. In addition, the IIRSA inherits some properties of ISRA such as convergence.

A frameless stereotactic localization method with DSA is introduced in this paper. A locating plate and four head marks are used in the method. Using two projection images of DSA, the three dimension coordinate of any point can be calculated referring to the reference locating coordinate system. The method is the theoretic basis of locating brain structure such as the blood vessel using DSA.

A decoupling generalized predictive control algorithm was presented in this paper. This algorithm was proposed to control the temperature-humidity system of a practical air-conditioning environmental chamber. By introducing absolute humidity as an intermediate variable, temperature and humidity of the system can be decoupled. Then the two decoupled systems are controlled by generalized predictive control (GPC) algorithm at different sample rates. Many simulation experiments demonstrated the effectiveness of the algorithm. Besides, decoupling of temperature and humidity will be extremely helpful to the application of GPC on the real system. This is also the effective method to control the temperature and humidity system.

Extended X-ray absorption fine structure (EXAFS) has been extensively used in the structural study in recent years. But the current methods of analysis EXAFS spectrum is difficult and tedious,demanding either very similar standard samples or the initial parameters near the solution. In this paper, a new technique Genetic Algorithms (GAs), was applied to the analyzing of the EXAFS spectrum,and two spectra of CU samples were investigated by the method. It was found that reasonable results can be obtained by the method only with limiting the initial parameters to a broad range. Comparing the results with those obtained by EXCURVE88, the fitting is superior, and the parameters are reasonable. Therefore, it is proved that the GAs may be a high performance method for the analyzing of the EXAFS spectra.

This paper presents a simple and quick algorithm, based on the principle of OR operation, for detecting macro-particles when the distribution of macro-particled is few and scattered. Counting and measuring the basic size parameters of macro-particles will be simultaneously accomplished by scanning the macroparticle image only once time.

ECG data compression has been greatly improved since 1960 s. Although the error of compression can be controlled accurately, the reported algorithm at present can not be processed in real-time with a microcomputer. In this paper an algorithm for ECG data compression is expounded which can be realized in real-time with a single chip computer. The algorithm has not only adjustable precision but also an average compression rate of 10.34, especially a pretty high compression ratio for frequent arrhythmic patients.

In this paper, linear programming and constrained simulated annealing are combined to optimize X-knife stereotactic radiotherapy treatment planning. In the first phase, linear programming reduces the resolution space, which is favourable to the adoption of little granularity in the second phase. Our experiment demonstrates that after the two-phase optimization, we attain the expected result.

Recently, heart rate variability (HRV) as a non-invasive measure for assessing the balance of autonomic nervous system has drawn more attentions and HRV analysis has become a focus of the ECG signal processing. QT interval represents the ventricular repolarization duration. QTV might reflect the balance of autonomic nervous system in other way. QTV analysis might complement HRV analysis. Because of the difficulty in detecting QT interval, RT interval is often used instead of QT interval. HRV and QTV detecting methods have been presented in this paper. After recording 24 h ECG signals, R fiducial marks are detected correctly and ectopic beats are removed. T peaks are detected on the basis of templet matching and correlation. The analysis in frequency domain consists of 24 h 3-dimensional frequency chart, as well as trends of HF/LF and HF, with fast Fourier transform and autoregressive techniques. We acquired ECG signals from subjects under controlled respiration in different frequencies. The results of HRV and RTV spectral analysis have shown that HF peak is modulated at the frequency of controlled respiration, which proves that HF peak is dependent on the vagal activity and our method is correct. The establishment of these methods has provided a basis for further research, and the development of HRV and QTV analysis techniques will be more helpful for early diagnosis, monitoring and prognostic assessment of some diseases.

Motor unit potentials(MUP) are real-time classified with the method of correlation in this paper. The correlation coefficients are sorted to reserve the better waves of MUP. Not only the store waves and analytical time of MUP can be reduced, but also the measurement correction and precision of MUP parameters can be improved.

This paper discusses the method of oscillometric blood pressure determination. A new criterion for systolic pressure and diastolic pressure determination has been proposed, which is based on ratio changes of the oscillation amplitude under a wide normalized range. Primary experiment results show the new criterion is effective.

Computerized Tomography needs projected images from all directions, tomosynthesis doesn't need the information from all directions, but its reconstruction image is indistinct. In this paper, we introduce Frequency Domain Deconvolution Imaging(FDDI) which proposed recently. FDDI overcomes shortcoming of tomosynthesis, but FDDI's calculation is large, and its zero frequency will lose. We propose a new method-image displacement elimination method(IDEM), it overcome shortcomings of FDDI and can solve convolution equation set in space domain quickly.

This paper addressed the application of wavelet to the measurement of QT dispersion and presented a new method using multi-thresholds to determinate the end point of T-wave. The results geined by computer and human were compared. We also suggested that QT dispersion be defined as the SD of 12 QT intervals.

This paper presents an analysis of Lyapunov exponents of ten normal persons and ten patients with diabetes mellitus by the definition of the Lyapunov exponents, the length evolution and the area evolution. The results show that the Lyapunov exponent of diabetic is clearly lower than that of normal. It can indicate the diabetic with less number of electrocardiographic data.

This paper explains the design of an Ultrasonics Doppler Spectrum Analysis System for fetal umbilical blood flow monitoring. The gist of discussion includes the sampling method of the Doppler signals and the algorithm of the blood flow parameters. Because the DSP chip is not used in the AD cardboard, the hardware design is simplified and the system works steadily. In the process of computing the blood flow paramenters, an autochange method of the amplitude threshold and a timeout control technology are used, so the algorithm has good adaptability and works very stable.

Based on the local gray level histogram of macro-particle image which is more special than other images, a new adaptive noise filtering method for macro-particle image processing is developed and described in this paper. It is made up of median filter and special order statistic (OS) filters (alpha-mean filter and inverse alpha-mean filter). This adaptive filter not only has satisfactory performance in both of noise-removal and edge-preservation of macro-particle image, but also is easy and quick to implement. An example is presented in this paper.

Based on the previous research about the feature of liver ultrasonic images, this paper presents a new analysis method of multifractal spectrum alpha-f(alpha). Using a series of partition Hausdorff dimension f(q) this paper describes the sophisticated fractal characteristics of the ultrasonic images in the different section of the growing process and the multi-fractal spectrum alpha-f(alpha) acts as the new parameter of the ultrasonic issue characterization. Using the neural network it has catalogued and recognized the tissue diseases, caring out ultrasonic tissue characterization. Apparently, this method has practical significance.

Considering the specificity of the medical color image, in this paper a medical color image is decomposed as three color orthogonal feature: I1, I2 and I3 in K-L transformation method, and Kirsch algorithm is used to detect the edge of the monochrome image I1 including rich information. Finally, both clarity edge and rich detail of color backdisplay image are gained, if we adopt I2 and I3 to compensate I1.

Fractal analysis is a new method for biomedical signal processing. Present research on surface EMG and its parameterization method were introduced in this paper. Experiments were performed on three conditions: (1) different muscle contraction degree, (2) different static load, (3) different motion pattern. Problems and prospects in engineering application were also discussed here.

Based on the criterion whether extracting the features of the image or not, the content and research status of medical image registration are discussed in this paper.

In this paper, a measuring system of pulse oxygen saturation is presented, which is based on the Lambert-beer Principle, using the relative intensities at two wavelengths in the red and infrared ranges. In order to meet the special demand of continuous clinical physiological monitoring, two circuits and appropriate algorithms are designed, one circuit, a feedback circuit controlled by micro-controller, automatically regulates the base line of signal, another circuit is a controlled integral circuit which automatically regulates the amplitude of waveforms. As the result, the drift of base line in this system, especially the effects caused by the physiological differences of human bodies, which will deteriorate the system, have been reduced. The system is satisfactory to clinical monitoring.

In this paper, we discuss the application of wavelet transform to edges detection, classification and description. Then, according to the peculiarity of the nail fold micro-circulation images, dyadic wavelet transform is successfully applied to detect the edges and the Mallat algorithm is used in the wavelet decomposition.

In this paper, a new algorithm based on Voxel Model is proposed. This algorithm combines the advantages of surface rendering and direct volume rendering. In the process of 3D reconstruction, OpenGL graphic standard and hardware accelerator can be used. By doing so, better reconstruction, result and fast speed can be achieved. Also, it is simple to turn this algorithm into programs.

A denoising algorithm for dynamic esophageal pH monitoring signal is first introduced in this paper. The algorithm was developed on the basis of multiresolution analysis. The pH monitoring signal was investigated in experiments, then the algorithm was formulated by multiresolution analysis based on discrete dyadic wavelet transform combining with the results of the investigation mentioned above. Application of this algorithm to clinical data processing has proved its satisfactory denoising effect.

This paper proposes the adaptive directional filter, which is exploited for the extracting of left ventricular(LV) contour. This algorithm could produce the exact ridge point by applying directional filter with smoothing operator in the angiography images damaged by noise the Edge Directional Vector as the result of this filter is efficient for predicting the edge ahead, while the filter parameters is adaptive. This proposed method provides the more precise and more automatic--extracting LV contour only with fewer interactions. The experimental results show that this proposed algorithm could improve the accuracy and reproducibility, and reduce the computational consumption remarkably.

This paper discusses the characteristic detection of modulus maxima of the loavelet transform, based on the properties of the singularity of the signal and noise. Theoretical analysis and experimental results show That the method is fitting for ECG signal detection and positioning.

It is an important topic in electroencephalography (EEG) research to localize the EEG activity sources from the scalp recordings. In this paper, based on the fourth-order cumulant matrix, a new sub-space decomposition algorithm is proposed for the EEG inverse problem. As the second-order moments (cumulants) has the drawback of being sensitive to the noise covariance. Using the fourth-order cumulants we need not know the noise covariances, as long as the noise is Gaussian. Computer simulation study on a three-layer concentric sphere head model shows its better performance than the two-order cumulate method in depressing the spatial coherent Gaussian noise.

Generally, each single action is picked up from successive action EMG signal by hand. Then this colleced signal is discriminated after getting its features. But doing by this way is obviously difficult to meet requirements of the myoelectric artificial limb. So this paper presents an easy method from the myoelectric energy point of view. The action signal segments and no-action signal segments are picked up from colleced EMG signal. Thus it is convenient for the relevant follow-up processing, and avoids dealing with no-action signal data. Therefore it is of useful and practical significance for multifunction myoelectric control of prostheses. In this paper, it has been proved that this method is feasible by using it to pick up actual effective action signal.

A wavelet packet-based algorithm for the compression of ECG is presented in the paper. The wavclet packet algorithm, which synthesizes the efficiency and the flexibility of the wavelet pacheket, generates significantly lower data rates with less computational effort.

The mechanism of forest self-thinning is generally nonlinear and dynamic, and the artificial neural network has the characteristic of expressing arbitrary nonlinear mapping. In this paper, the feasibility and limitation of artificial neural network used to simulating forest self-thinning was expounded, and the principle and algorithms of the neural network model based on modified simplex method (BP-MSM mixed algorithms) for modeling forest self-thinning were described. Its applications in self-thinning of Populus tremula natural forest Cunninghamia lanceolata plantation were illustrated. The results of forest self-thinning examples show the BP-MSM mixed algorithms were satisfactory in simulating forest self-thinning, and its precision was higher, which develops the method and theory of artificial neural network, and enriches the simulating method of forest self-thinning.

The theory of autocorrelation algorithm for color flow mapping is analyzed in detail. The digital accomplishment for the algorithm in this paper is also analyzed. With the consistent theory bases, high precision and simply realization, the technology will lay a good foundation for its medical applications in hospitals.

A designing method for chromosome automatic analyzing system is presented in the paper. The system accomplishes chromosome automatic recognizing, sorting, pairing, picking out aberrant chromosomes and then makes diseases-diagnosis.

This paper presents a new medical data visualization method called multi-resolution blending rendering which has been developed for visualizing the large medical volume data sets. Different from those multi-resolution rendering methods which simplify the structures by polygon mesh optimization technique, the new method uses different resolution levels to represent different parts of a structure at the same time. So, it can preserve the integrality of the structure and emphasize the important part of the structure when simplifying the structure. The new method also takes advantage of the good qualities of the wavelet transform and constructs the multi-resolution blending model of the structure by orthogonal wavelet transform and transition operator. The experiment proves that our method can effectively simplify the model of the structure, enormously reduce the number of the triangle of the structure and speed up the structure rendering. It is especially suitable for interactive observations on a complex medical structure.

In this paper, we discuss the relation between matched filter and wavelet transform(WT), and point out that wavelet transform is just the matched filter with changeable detection template. According to this idea, the method of signal detection based on hybrid wavelet transform (HWT) is proposed. HWT in this paper means that in WT decomposition and reconstruction, we use two different mother wavelets. One is used as a changeable template for the pulses detection and the other is used for the characteristic enhancement of detected pulse. This method has been applied to the interference pulse detection in EEG signal. The experiment result shows that HWT has the good property for instantaneous signal detection.

Edge-detection is all the time a major problem in the computer early vision, and it plays an important role in image processing. This paper reviews classical and new methods of edge-detection and discusses its application in medical image processing.

Through mechanical analysis, it is discovered that the twined structure decides the amplification coefficient of the catgut tension. Many troubles of production can be explained with this mechanics model, and it could be used as a reference for medical sutures. The paper gives an accurate calculating formula for catgut design and its direct tensile strength through improving mechanics model. It can raise the material strength from 10 to over 31% depending on its twined structure.

Contour detection is one of the most difficult problems in the multi-dimensional reconstruction of echocardiographic images. A method based on active contour model is presented in the present paper to solve this problem. First, according to the characteristic of the ultrasonic medical images, an adaptive weighted median filter is used to suppress speckle noise and an adaptive threshold value is selected to get binary image. Then, mathematical morphological treatment is utilized to get the initial contour. At last, active contour model is employed to revise the initial contour and get the accurate final contour. Experiments with real and interpolated ultrasonic images are presented. The results show the effectiveness of this method. This method has practical use in the multi-dimensional reconstruction of ultrasonic medical images.

In magnetoencephalogram(MEG) basic studies, it is an important issue to estimate magnetic source parameters by inverse solution. It is known that the magnetic field equations are nonlinear, thus explicit solutions are difficult to obtain. However optimization methods are available to this parameter estimation. In many usually used nonlinear local optimization algorithms, Gauss-Newton's is of fast convergent speed. When this algorithm is used, the singularity of the Jacobien matrix about the minimum least square error must be considered carefully. If the matrix is singular, the equation for searching direction has no general solution. One way to overcome this problem is to use negative gradient as searching direction, but it may cause descent of convergent speed. Another way is known as Levenberg-Marquardt algorithm which makes the matrix non-singular by adding some improved factors to it. In this paper we utilize Moore-Penrose inversion for the solution of iterative searching direction equation. In appendix we demonstrate that the searching direction obtained by the proposed method is successful. Computer simulation also demonstrates that by reasonable selection of initial iterative values, the modified Gauss-Newton algorithm is effective for MEG inverse solution in the case with one or two source dipoles.

This paper presents a discussion on the tracing algorithms of the extraction of Electrocardiogram (ECG) data from scanned image of hardcopy ECG. We have advanced a new algorithm integrated by seek-and-predict method and polygon model method, which can tolerate crossing of two leads, fault-line and disturbing of grid lines. Two practical situations that may be encountered while processing hardcopy ECG are discussed in detail. Furthermore, the after-treatment of extraction-filtering is considered and an SVD-based filter designed to smooth ECG signal is analyzed.

The existing algorithms for displaying the cutting isosurface of the structure are not suitable for interactive display because they spend too much time on calculation when a fair amount of data is submitted to processing. This paper presents a new algorithm for accelerating the cutting of structure by using the area partition and the increment calculation to reduce the unnecessary calculation in cutting the structure. It is suitable for real time interactive displaying, especially in the case of a vast amount of data.

This paper, utilizing signal power spectrum and L-Z complexity measure, analyzes the abnormal ECG: ventricular tachycardia (VT) and ventricular fibrillation (VF). And on the base of linear and nonlinear dynamics, the paper forwards a new definition of information complexity rate, deduces some relative properties and applies it to complexity analysis in irregular ECG. By means of biological experiments and computer simulations, the reasonability of information complexity and complexity rate is confirmed. At last, objective analysis and explanations of the mechanism of VT and VF are reported. The results indicate: with the help of power spectrum, complexity measure and complexity rate, the recognition of VT and VF can attain 100%.

In this paper, we present a PID algorithm with velocity project of voice coil motor controlled on the base of H8/338MCU, and we find a driving part, which it is smaller, lighter and more reliable, for the left ventricular assist device. By means of the simulated experiment on the system, the result is satisfying. So we can say that it is a good help for the realization of LVAD.

A design of a Digital Scan Converter(DSC) in B scanner and its implementation by using FPGA are presented in this paper, and some applied results are presented too here.

It is very important to monitor the fetal heart rate in the clinical management of pregnancies antenatally. The most commonly employed non-invasive method uses Doppler ultrasound. Because of the complexity of Doppler fetal signals, the signal processing is difficult. In this paper, we describe an improved method for the determination of fetal heart rate from Doppler ultrasound signals. Using an exponentially weighted auto-correlation technique with an automatic adaptation algorithm, we have developed the method which can provide accurate fetal heart rate records, especially in the cases of reduced signal and loss of signal.

In this paper, a new method is presented to detect the edge of a colonscope image using cublic B-spline wavelet. This method can make up for the shortcoming of traditional method, resulting from blue dark area in colonscope image through estimating the circle centers. These estimated circle centers which have arc-shaped edge lines are weighted to obtained the lumen centers. This center is used to navigate colon endoscope in colonscopic operation.

The liver moves in response to impulses arising from the main blood vessels. The hardness of the liver may be reflected by this kind of movement. The M-mode ultrasound technique is used in this paper to detect the information of livers' motion. The motion envelope of the liver excited by the abdominal aorta is extracted using the correlation technique. Several feature parameters, such as the amplitude, the standard deviation and Fourier harmonics, are extracted from this envelope. The values of parameters were calculated from 19 normal livers and 11 patients' livers, from which the criterion was obtained to decide whether the liver was normal.

An algorithm based on wavelet transforms (WT'S) has been developed for detecting QRS complexes of ECG signal. In this paper, we use Mallat algorithm to calculate dyadic Marr wavelet transforms, analyse the relationship between singularity of the signal (the peak of R wave) and local maximal value of WT's from the view of equivalent wavelet filter, discuss the performance distinction between quadratic differential wavelet and single differential wavelet for singularity detection. Some other tactics are also applied to improve the anti-interference performance of the algorithm and to increase the correct detection rate of QRS complexes. Verified by the MIT-BIH Arrhythmia Database, the correct detection rate of QRS complexes is 99.8%.

This paper introduces an ECG computing algorithm that computes ECG based on the action potential of single cardiac cell. Taking solid angle analysis as tool, this algorithm analyzes the field potential of a cardiac cell that connects and communicates with neighboring cells. This algorithm has been implemented in our electrophysiological model CardioAuto, which was built with extended Cellular 3.0-a cellular automata system. Algorithm and simulation results reveal that it is the transmembrane potential slope among cell group that contributes to ECG waveform generation and determines the waveform deflection. To a specific lead location, whenever cell group has a transmembrane potential slope that proximal end is higher than distal end, a downward ECG deflection is generated. Whenever cell group has a reverse transmembrane potential slope, an upward ECG deflection is generated. Whenever there is no transmembrane potential slope, ECG keeps on baseline. According to our algorithm, the significance of normal and many abnormal ECG waveforms can be analyzed and interpreted at cellular level.

This paper introduces the backgrounds, principles and some popular algorithms used for resolving calibration transfer in multivariate calibration of chemometrics. The algorithms include S/B, DS, PDS, FIR, and Shenk's algorithm. The recent development of calibration transfer in the past two years is also reviewed. A number of key applications of different calibration transfer algorithms are summarized, which will be helpful for the further research of calibration transfer principles and algorithms. 51 references are cited.