Catálogo de publicaciones - libros

Method for extracting the specified rhythms of clinical electroencephalogram (EEG) is proposed using the wavelet packet decomposition. Based on the ability of accurately resolving the signal into desired time-frequency components, EEG signals are preprocessed and decomposed into a series of rhythms for many clinical applications. Specified dynamic EEG rhythms can be accurately filtered with designed wavelet structure. In addition, we present a wavelet packet entropy method for processing of EEG signal. Both relative wavelet packet energy and wavelet packet entropy are presented as the quantitative parameter to measure the complexity of the EEG signal. Several experiments with real EEG signals are carried out to show that the proposed method excels the common discrete wavelet decomposition. The presented procedure can isolate specific EEG rhythms accurately and is also regarded as an efficient method for analyzing non-stationary signals in practice.

The relative performances of different classifiers applied to the same data are typically analyzed using the Receiver Operator Characteristic framework (ROC). This paper proposes a further analysis by explaining the operation of classifiers using low-order Boolean rules to fit the predicted response surfaces using the Orthogonal Search Based Rule Extraction algorithm (OSRE). Four classifiers of malignant or benign ovarian tumours are considered. The models analyzed are two Logistic Regression models and two Multi-Layer Perceptrons with Automatic Relevance Determination (MLP-ARD) each applied to a specific alternative covariate subset. While all models have comparable classification rates by Area Under ROC (AUC) the classification varies for individual cases and so do the resulting explanatory rules. Two sets of clinically plausible rules are obtained which account for over one half of the malignancy cases, with near-perfect specificity. These rules are simple, explicit and can be prospectively validated in future studies.

Air-pressure limit value is an important conditional parameter of artificial respiration. The pulmonary characteristics are very different according to the person. For setting appropriate ventilation conditions fitting to each patient, it is necessary to establish a mathematical model describing the mechanism of human respiratory system, and to know the pulmonary characteristic of each patient via identification of the model. For this purpose, two types of respiratory system models have been proposed by the authors. These models are expressed as second order nonlinear differential equations with air-volume variant elastic coefficient and air-volume variant resistive coefficient. In the first type of model, elastic coefficient is expressed as polynomial function of air-volume, while in the second type of model, elastic coefficient is expressed by RBF network. The model with polynomial expression of elastance has the advantage that the structure is simple. On the other hand, the model with RBF expression of elastance has better numerical stability against to the model with polynomial expression of elastance. In this paper, a decision method of air-pressure limit value based on the respiration model with RBF expression of elastance is proposed. This method adopt a numerical technique to find the point of saturation starting point in the elastance curve, So direct calculation of radius of curvature can be avoided. The proposed method is validated by an example of application to practical clinical data.

Micorarray data are often extremely asymmetric in dimensionality, such as thousands or even tens of thousands of genes and a few hundreds of samples. Such extreme asymmetry between the dimensionality of genes and samples presents several challenges to conventional clustering and classification methods. In this paper, a novel ensemble method based on correlation analysis is proposed. Firstly, in order to extract useful features and reduce dimensionality, different feature selection methods based on correlation analysis are used to form different feature subsets. Then a pool of candidate base classifiers is generated to learn the subsets which are re-sampling from the different feature subsets. At last, appropriate classifiers are selected to construct the classification committee using EDA (Estimation of Distribution Algorithms) algorithm. Experiments show that the proposed method produces the best recognition rates on two benchmark databases.

The detection of explosives and illicit material in passengers’ luggage for the purpose of station security is an important area in public traffic security. This paper presents a method for X-ray image recognition based on fuzzy rule and parallel neural networks. Neural networks have been widely used in various fields. However, the computing efficiency decreases rapidly if the scale of neural network increases. In this paper, a new method of X-ray image recognition based on the fuzzy-neuron system is proposed. In fuzzy rules method, a test pattern may belong to several classes with different degrees. A neural networks classifier is just for one class and used to make sure if the pattern is really belonged to that class based on fuzzy rules, they are combined to obtain the recognition result. From the experience results, the new method performs well.

In prediction of secondary structure of proteins there are always some suspected segments. These suspected segments confuse people and lower the accuracy of prediction methods. To deal with this problem, a set of neural networks (NNs) are built based on helix, strand and coil segments selected from PDB. The test performance of these NNs on training data is perfect without surprise. However the prediction on test data is not good enough because the training data are lake of great representativeness. The results support the fact that closer neighbor vectors have the similar outputs of NNs. One can improve representativeness of training data without enlarging data scale as long as select less data from dense region and more from sparse region on condition that distribution of sample data has been known.

Detecting the boundaries of protein domains has been an important and challenging problem in experimental and computational structural biology. In this paper the domain detection is first taken as an imbalanced data learning problem. A novel undersampling method using distance-based maximal entropy in the feature space of SVMs is proposed. On multiple sequence alignments that are derived from a database search, multiple measures are defined to quantify the domain information content of each position along the sequence. The overall accuracy is about 87% together with high sensitivity and specificity. Simulation results demonstrate that the utility of the method can help not only in predicting the complete 3D structure of a protein but also in the machine learning system on general imbalanced datasets.

In [1], we developed two methods to automatically identify the contribution of the recording reference signal from multi-channel intracranial Electroencephalography (iEEG) recordings. In this study, we subtract the reference recording contribution to iEEG and obtain corrected iEEG. We then investigate three commonly used iEEG metrics: spectral power, phase synchrony, and magnitude squared coherence (MSC) for common referential iEEG, corrected iEEG and bipolar montage iEEG. We find significant differences among the three iEEG metrics, and are able to determine the contribution from the recording reference to each metric. Generally, reference signals with smaller amplitude yield lower phase synchrony and reference signals with larger amplitude increase phase synchrony. Reference signals with spectral peaks increase coherence. Reference signal with low power may have no significant impact on calculated coherence. Bipolar EEG usually yields small phase synchrony or MSC values and may obscure the actual phase synchrony or MSC values between two local sources.

Shape description is the precondition for shape matching and retrieval. The more robust and stable primitives to describe shapes are global topological properties, but obtaining global topological properties is still an obstacle in computer vision. Motivated by the difference sensitivity of short-range connection in biology vision, we present a novel global descriptor to describe the entire topology of simple closed 2D shape in this paper. We employ two novel strategies – the zigzag rule, which approximates shape to an elaborate polygonal curve, and cost function which combines global configurations as well as local information of the line stimulations as our punishments. With these two key steps the descriptor is robust to translation, scaling and rotation. Experimental results show the model gain good performance on matching and retrieval for silhouettes. Even for images with occlusion the result is excellent and reasonable.

For the optional control to the trans-femoral prosthesis and natural gait, an ongoing investigation of lower limb prosthesis model with myoelectrical control was presented. In this research, the surface electromyographic signals of lower limb were extracted to be switch signal, and translate into movement information. Considering every muscle’s different physiologic tendency, fuzzy support vector regression method was applied to establish an intelligent black box that can interpret the physiological signals to accurate information of knee joint angle. It achieves a comparable or better performance than other methods, and provides a more native gait to the prosthesis user.