Using a Combination of Model Based and Intelligent methods in Automatic Landmark Detection in Cephalometry

This paper introduces a modification on using active shape models (ASM) for automatic landmark detection in cephalometry and combines many new ideas to improve its performance. In first step, some feature points are extracted to model the size, rotation, and translation of skull. A learning vector quantization (LVQ) neural network is used to classify images according to their geometrical specifications. Using LVQ for every new image, the possible coordinates of landmarks are estimated, knowing the class of new image. Then a modified ASM with a multi resolution approach is applied and a principal component analysis (PCA) is incorporated to analyze each template and the mean shape is calculated. The local search to find the best match to the intensity profile is then used and every point is moved to get the best location. Finally a sub image matching procedure, based on cross correlation, is applied to pinpoint the exact location of each landmark after the template has converged. On average It percent of the landmarks are within 1 mm of correct coordinates,percent within 1 mm, and percent within 1 mm, which shows a distinct improvement on other proposed methods.

Diabetes Mellitus is a chronic, non-infectious disease that affects people’s health, which has a rapid onset trend with the continuous improvement of China’s economy and material life. At present, there is no practical solution to this condition. Diabetic retinopathy (DR) is the most important manifestation of diabetic microangiopathy, which can be divided into proliferative lesion and non-proliferative lesion. The traditional artificial diagnosis method has strong subjectivity and low accuracy. Because disease diagnosis can be regarded as a two-classification pattern recognition problem, the application of neural network method can provide the possibility for the application of artificial intelligence (AI) in disease-assisted diagnosis and treatment. Furthermore, it has significant meanings to find which kind of neural network has a better efficiency. In this paper, learning vector quantization (LVQ) neural network and back propagation (BP) neural network were used to diagnose diabetes and diabetic retinopathy with MATLAB and their recognition rate were compared. The datasets of diabetes and diabetic retinopathy were available in the UCI database. The experiment results were analyzed to evaluate the efficiency of each neural network classifier. The results demonstrate that both LVQ neural network and BP neural network can classify the two datasets effectively. However, compared with the LVQ neural network, the average accuracy rate and sensitivity of the BP neural network is higher.

The Raman spectra of tissue of 20 brain tumor patients was recorded using a confocal microlaser Raman spectroscope with 785 nm excitation in vitro. A total of 133 spectra were investigated. Spectra peaks from normal white matter tissue and tumor tissue were analyzed. Algorithms, such as principal component analysis, linear discriminant analysis, and the support vector machine, are commonly used to analyze spectral data. However, in this study, we employed the learning vector quantization (LVQ) neural network, which is typically used for pattern recognition. By applying the proposed method, a normal diagnosis accuracy of 85.7% and a glioma diagnosis accuracy of 89.5% were achieved. The LVQ neural network is a recent approach to excavating Raman spectra information. Moreover, it is fast and convenient, does not require the spectra peak counterpart, and achieves a relatively high accuracy. It can be used in brain tumor prognostics and in helping to optimize the cutting margins of gliomas.

Standard back propagation (BP) neural network has disadvantages such as slow convergence speed, local minimum and difficulty in definition of network structure. In this paper, an learning vector quantization (LVQ) neural network classifier is established, then it is applied in pattern classification of two-dimensional vectors on a plane. To test its classification ability, the classification results of LVQ neural network and BP neural network are compared with each other. The simulation result shows that compared with classification method based on BP neural network, the one based on LVQ neural network has a shorter learning time. Besides, its requirements for learning samples and the number of competing layers are also lower. Therefore it is an effective classification method which is powerful in classification of two-dimensional vectors on a plane.

The design of pattern classifiers such as multiprototype classifiers and neural network classifiers such as learning vector quantization and radial basis function neural networks requires reducing the size of the training data sets. In addition, memory storage, computation complexity and time, and data redundancy demand many pattern classifiers to use a smaller subset of a training data set. In this paper, we present a data reduction algorithm which automatically selects the subset of training data that faithfully represents the training data set for pattern classification. The applicability of this algorithm is demonstrated through k-nearest neighbor and learning vector quantization neural networks classifiers using both speech and synthetic data sets.

A learning vector quantization(LVQ) neural network classifier is for the first time applied for SAR data classification, of which the training termination rule is modified to make it have both the ability of classification and class signature analysis. A very high land cover classification accuracy is achieved. Especially under the condition that texture is considered, almost all roads can be classified correctly, which cannot be identified by BP MLP neural network and ML classifier.

Transformer inrush currents are high magnitude, harmonic-rich currents generated when transformer cores are driven into saturation during energization. In this paper an efficient method for detection of inrush current in distribution transformer based on wavelet transform is presented. Using this method inrush current can be discriminate from other transients such as capacitor switching, load switching and single phase to ground fault. Wavelet transform is used for decomposition of signals and Learning Vector Quantizer(LVQ) neural network used for classification. Inrush current data and other transients are obtained by simulation using EMTP program. Results show that the proposed procedure is efficient in identifying inrush current from other events.

Tractor condition recognition has important research value in helping to understand the operating status of tractors and the trend of tillage depth changes in the field. Therefore, this article presents a method for recognizing tractor conditions, providing the basis for establishing the relationship between tractor conditions and the tillage depth of the attached agricultural machinery. This study designed a tractor condition recognition method based on neural networks. Using real-world vehicle data to establish a data set, K-means clustering analysis was used to label the data set based on four conditions: “accelerated start”, “constant speed”, “decelerated stop” and “turning”. The learning vector quantization (LVQ) neural network and the VGG-16 model of a CNN were selected for use recognizing the tractor conditions. The results showed that both the neural networks had good recognition effects. The average accuracy rates of the VGG-16 model of CNN and LVQ neural network were 90.25% and 79.7%, respectively, indicating that these models could be applied to tractor condition recognition and provide theoretical support for the correction of angle detection errors.

It is an important part in aluminum electrolysis production to control the anode effect (AE). Since there are some shortcomings in traditional methods of anode effect prediction in aluminum electrolysis, this paper combined two methods, the slope of cell resistance and learning vector quantization (LVQ) neural network, to predict anode effect. First of all, the first prediction of anode effect will be conducted based on the slope of cell resistance. Afterwards, the inaccurate data are supposed to be re-predicted. The second prediction consists of two steps, one is to estimate the power spectrum from the signal of cell resistance by means of periodogram, the other is to re-predict the anode effect with the LVQ neural network, since the energy of frequency bands are served as the input feature variables of neural network, so as to raise the accuracy of prediction. It turned out that the success rate of ten-minute in advance prediction for anode effect can be above 85%, though just cell resistance signal is studied.

This paper proposes a framework for developing a broad variety of soft clustering and learning vector quantization (LVQ) algorithms based on gradient descent minimization of a reformulation function. According to the proposed axiomatic approach to learning vector quantization, the development of specific algorithms reduces to the selection of a generator function. A linear generator function leads to the fuzzy c-means (FCM) and fuzzy LVQ (FLVQ) algorithms while an exponential generator function leads to entropy constrained fuzzy clustering (ECFC) and entropy constrained LVQ (ECLVQ) algorithms. The reformulation of clustering and LVQ algorithms is also extended to supervised learning models through an axiomatic approach proposed for reformulating radial basis function (RBF) neural networks. This approach results in a broad variety of admissible RBF models, while the form of the radial basis functions is determined by a generator function. This paper shows that gradient descent learning makes reformulated RBF neural networks an attractive alternative to conventional feed-forward neural networks.

We present a PCA-LVQ model and a balanced-training method for efficient intrusion alert analysis. For the connection records in the 1999 DARPA intrusion dataset, we firstly get a dimension-reduced dataset through Principal Component Analysis (PCA). Then, we use the Learning Vector Quantization (LVQ) neural network to perform intrusion alert clustering on the purified intrusion dataset. The experiment results show that the PCA-LVQ model and the balanced-training method are effective: the time costs can be shortened about by three times, and the accuracy of detection can be elevated to a higher level, especially for the U2R and R2L alerts.

Timely identification of the aging stage of the transformer can effectively ensure the safe operation of the transformer and prevent the aging fault of the transformer. This work proposes the aging diagnosis of transformer using the Raman spectroscopic recurrence plot combined with learning vector quantization (LVQ) neural network. Oil-paper insulation samples at different aging stages are obtained by accelerated thermal aging tests. Acquisition of Raman spectra of insulating oil is from 800 to 3000 cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−1</sup> and converts to the Raman spectroscopic recurrence plot. The gray-level cooccurrence matrix (GLCM) is used to extract the characteristic information of the Raman spectroscopic recurrence plot. The optimal number of neurons in competitive layer of LVQ neural network is determined by the experimental method. A transformer aging diagnosis model based on the Raman spectroscopic recurrence plot is established. The results show that the diagnostic accuracy of the proposed diagnostic model reached 95%. The Raman spectroscopic recurrence plot has more abundant aging “fingerprint” characteristic information and the proposed diagnostic method shows potential for monitoring oil-paper insulation.

The current study presents the hard implementation methodology of a Learning Vector Quantization (LVQ) neural network on a Field Programmable Gate Array (FPGA) circuit specially suited for fast output applications. The implementation methodology is based on a mixed parallel sequential approach with the use of the L2 norm (Euclidian distance) to measure the distance between the reference vector and the prototype vector. The adopted architecture has been implemented on a device XCV1000 (FPGA Xilinx) and the given results have shown good performances in time, surface and consumption.

The self-potential anomalies produced by simple polarized geologic structures are used in the position detection of buried objects such as rocks or minerals. If these objects are adjacent, a mixed self-potential anomaly data will be measured. However, the detection of the objects position from this mixed self-potential anomaly data is usually not possible. In this paper, the mixed self-potential anomaly data is first separated by a blind signal separation technique called the independent component analysis (ICA), then the learning vector quantization (LVQ) neural network is used in the position detection of the separated self-potential anomalies. The proposed system achieves very high accuracy

Video compression has great significance in the communication of motion pictures. Video compression techniques try to remove the different types of redundancy within or between video sequences. In the temporal domain, the video compression techniques remove the redundancies between the highly correlated consequence frames of the video. In the spatial domain, the video compression techniques remove the redundancies between the highly correlated consequence pixels (samples) in the same frame. Evolving neural-networks based video coding research efforts are focused on improving existing video codecs by performing better predictions that are incorporated within the same codec framework or holistic methods of end-to-end video compression schemes. Current neural network-based video compression adapts static codebook to achieve compression that leads to learning inability from new samples. This paper proposes a modified video compression model that adapts the genetic algorithm to build an optimal codebook for adaptive vector quantization that is used as an activation function inside the neural network’s hidden layer. Background subtraction algorithm is employed to extract motion objects within frames to generate the context-based initial codebook. Furthermore, Differential Pulse Code Modulation (DPCM) is utilized for lossless compression of significant wavelet coefficients; whereas low energy coefficients are lossy compressed using Learning Vector Quantization (LVQ) neural networks. Finally, Run Length Encoding (RLE) is engaged to encode the quantized coefficients to achieve a higher compression ratio. Experiments have proven the system’s ability to achieve higher compression ratio with acceptable efficiency measured by PSNR.