This paper presents two novel approaches to determine optimum growing multi-experts network (GMN) structure. The first method called direct method deals with expertise domain and levels in connection with local experts. The growing neural gas (GNG) algorithm is used to cluster the local experts. The concept of error distribution is used to apportion error among the local experts. After reaching the specified size of the network, redundant experts removal algorithm is invoked to prune the size of the network based on the ranking of the experts. However, GMN is not ergonomic due to too many network control parameters. Therefore, a self-regulating GMN (SGMN) algorithm is proposed. SGMN adopts self-adaptive learning rates for gradient-descent learning rules. In addition, SGMN adopts a more rigorous clustering method called fully self-organized simplified adaptive resonance theory in a modified form. Experimental results show SGMN obtains comparative or even better performance than GMN in four benchmark examples, with reduced sensitivity to learning parameters setting. Moreover, both GMN and SGMN outperform the other neural networks and statistical models. The efficacy of SGMN is further justified in three industrial applications and a control problem. It provides consistent results besides holding out a profound potential and promise for building a novel type of nonlinear model consisting of several local linear models.
In this paper, we investigate the use of a self-adaptive Pareto evolutionary multi-objective optimization (EMO) approach for evolving the controllers of virtual embodied organisms. The objective of this paper is to demonstrate the trade-off between quality of solutions and computational cost. We show empirically that evolving controllers using the proposed algorithm incurs significantly less computational cost when compared to a self-adaptive weighted sum EMO algorithm, a self-adaptive single-objective evolutionary algorithm (EA) and a hand-tuned Pareto EMO algorithm. The main contribution of the self-adaptive Pareto EMO approach is its ability to produce sufficiently good controllers with different locomotion capabilities in a single run, thereby reducing the evolutionary computational cost and allowing the designer to explore the space of good solutions simultaneously. Our results also show that self-adaptation was found to be highly beneficial in reducing redundancy when compared against the other algorithms. Moreover, it was also shown that genetic diversity was being maintained naturally by virtue of the system's inherent multi-objectivity.
RADARSAT data have a potential role for coastal pollution monitoring. This study presents a new approach to detect and forecast oil slick trajectory movements. The oil slick trajectory movements is based on the tidal current effects and Fay's algorithm for oil slick spreading mechanisms. The oil spill trajectory model contains the integration between Doppler frequency shift model and Lagrangian model. Doppler frequency shift model implemented to simulate tidal current pattern from RADARSAT data while the Lagrangian model used to predict oil spill spreading pattern. The classical Fay's algorithm was implemented with the two models to simulate the oil spill trajectory movements. The study shows that the slick lengths are effected by tidal current V component with maximum velocity of 1.4 m/s. This indicates that oil slick trajectory path is moved towards the north direction. The oil slick parcels are accumulated along the coastline after 48 h. The analysis indicated that tidal current V components were the dominant forcing for oil slick spreading.
A hybrid neural network comprising Fuzzy ARTMAP and Fuzzy C-Means Clustering is proposed for pattern classification with incomplete training and test data. Two benchmark problems and a real medical pattern classification task are employed to evaluate the effectiveness of the hybrid network. The results are analyzed and compared with those from other methods.
Legendre moments are continuous moments, hence, when applied to discrete-space images, numerical approximation is involved and error occurs. This paper proposes a method to compute the exact values of the moments by mathematically integrating the Legendre polynomials over the corresponding intervals of the image pixels. Experimental results show that the values obtained match those calculated theoretically, and the image reconstructed from these moments have lower error than that of the conventional methods for the same order. Although the same set of exact Legendre moments can be obtained indirectly from the set of geometric moments, the computation time taken is much longer than the proposed method.
Different approaches to gray and white matter measurements in magnetic resonance imaging (MRI) have been studied. For clinical use, the estimated values must be reliable and accurate when, unfortunately, many techniques fail on these criteria in an unrestricted clinical environment. A recent method for tissue clusterization in MRI analysis has the advantage of great simplicity, and it takes the account of partial volume effects. In this study, we will evaluate the intensity of MR sequences known as T1-weighted images in an axial sliced section. Intensity group clustering algorithms are proposed to achieve further diagnosis for brain MRI, which has been hardly studied. Subjective study has been suggested to evaluate the clustering group intensity in order to obtain the best diagnosis as well as better detection for the suspected cases. This technique makes use of image tissue biases of intensity value pixels to provide 2 regions of interest as techniques. Moreover, the original mathematic solution could still be used with a specific set of modern sequences. There are many advantages to generalize the solution, which give far more scope for application and greater accuracy.
Kinetics of production of biodiesel by enzymatic methanolysis of vegetable oils using lipase has been investigated. A mathematical model taking into account the mechanism of the methanolysis reaction starting from the vegetable oil as substrate, rather than the free fatty acids, has been developed. The kinetic parameters were estimated by fitting the experimental data of the enzymatic reaction of sunflower oil by two types of lipases, namely, Rhizomucor miehei lipase (RM) immobilized on ion-exchange resins and Thermomyces lanuginosa lipase (TL) immobilized on silica gel. There was a good agreement between the experimental results of the initial rate of reaction and those predicted by the proposed model equations, for both enzymes. From the proposed model equations, the regions where the effect of alcohol inhibition fades, at different substrate concentrations, were identified. The proposed model equation can be used to predict the rate of methanolysis of vegetable oils in a batch or a continuous reactor and to determine the optimal conditions for biodiesel production.
Artificial neural network (ANN) was used in this study to determine water turbidity by using back propagation algorithm. Three wavelengths which represent reflectance intensity for eight standard samples were used as training input. The finding from the study shows that the trained network with number of epochs of 250,000 and learning rate of 0.001 gave the lowest sum of squared error (SSE) of 0.04. ANN was able to predict the turbidity of water based on the pattern recognition of the reflectance spectrum. The architecture of optimized ANN used in this study was 3:25:1. The average prediction error was 0.02.
[Jaringan neural tiruan (ANN) dengan lagoritma perambatan balik (BP) telah digunakan dalam kajian ini untuk menentukan kekeruhan air. Tiga panjang gelombang yang mewakili serapan bagi lapan sampel telah dipilih sebagai imput latihan. Hasil kajian menunjukkan bagi jaringan terlatih dengan bilangan ulangan latihan 250,000 dan kadar pembelajaran 0.001 telah memberikan nilai SSE yang terendah iaitu 0.04. Dalam kajian ini jaringan ANN didapati boleh menentu dan meramalkan nilai kekeruhan sample air berdasarkan corak serapan pantulan. Arkitektur yang sesuai bagi kajian ini adalah 3:25:1. Purata ralat ramalan adalah 0.02].
High performance sorbents for flue gas desulfurization can be synthesized by hydration of coal fly ash, calcium sulfate, and calcium oxide. In general, higher desulfurization activity correlates with higher sorbent surface area. Consequently, a major aim in sorbent synthesis is to maximize the sorbent surface area by optimizing the hydration conditions. This work presents an integrated modeling and optimization approach to sorbent synthesis based on statistical experimental design and two artificial intelligence techniques: neural network and genetic algorithm. In the first step of the approach, the main and interactive effects of three hydration variables on sorbent surface area were evaluated using a full factorial design. The hydration variables of interest to this study were hydration time, amount of coal fly ash, and amount of calcium sulfate and the levels investigated were 4-32 h, 5-15 g, and 0-12 g, respectively. In the second step, a neural network was used to model the relationship between the three hydration variables and the sorbent surface area. A genetic algorithm was used in the last step to optimize the input space of the resulting neural network model. According to this integrated modeling and optimization approach, an optimum sorbent surface area of 62.2m(2)g(-1) could be obtained by mixing 13.1g of coal fly ash and 5.5 g of calcium sulfate in a hydration process containing 100ml of water and 5 g of calcium oxide for a fixed hydration time of 10 h.
In this paper, an algorithm based on independent component analysis (ICA) for extracting the fetal heart rate (FHR) from maternal abdominal electrodes is presented. Three abdominal ECG channels are used to extract the FHR in three steps: first preprocessing procedures such as DC cancellation and low-pass filtering are applied to remove noise. Then the algorithm for multiple unknown source extraction (AMUSE) algorithm is fed to extract the sources from the observation signals include fetal ECG (FECG). Finally, FHR is extracted from FECG. The method is shown to be capable of completely revealing FECG R-peaks from observation leads even with a SNR=-200dB using semi-synthetic data.
This study presents the development of a hybrid system consisting of an ensemble of Extended Kalman Filter (EKF) based Multi Layer Perceptron Network (MLPN) and a one-pass learning Fuzzy Inference System using Look-up Table Scheme for the recognition of electrocardiogram (ECG) signals. This system can distinguish various types of abnormal ECG signals such as Ventricular Premature Cycle (VPC), T wave inversion (TINV), ST segment depression (STDP), and Supraventricular Tachycardia (SVT) from normal sinus rhythm (NSR) ECG signal.
Pulmonary atresia with intact ventricular septum (PAIVS) is a disease with remarkable morphologic variability, affecting not only the pulmonary valve but also the tricuspid valve, the RV cavity and coronary arteries. With advances in interventional techniques and congenital heart surgery, the management of PAIVS continues to evolve. This review is an attempt at providing a practical approach to the management of this disease. The basis of our approach is morphologic classification as derived from echocardiography and angiography. Group A, patients with good sized RV and membranous atresia, the primary procedure at presentation is radiofrequency (RF) valvotomy. Often it is the only procedure required in this group with the most favourable outcome. Patients with severely hypoplastic RV (Group C) are managed along the lines of hearts with single ventricle physiology. The treatment at presentation is patent ductus arteriosus (PDA) stenting with balloon atrial septostomy or conventional modified Blalock Taussig (BT) shunt. Bidirectional Glenn shunt may be done 6-12 months later followed by Fontan completion after a suitable interval. Patients in Group B, the intermediate group, are those with borderline RV size, usually with attenuated trabecular component but well developed infundibulum. The treatment at presentation is RF valvotomy and PDA stenting +/- balloon atrial septostomy. Surgical re-interventions are not uncommonly required viz. bidirectional Glenn shunt when the RV fails to grow adequately (11/2 - ventricle repair) and right ventricular outflow tract (RVOT) reconstruction for subvalvar obstruction or small pulmonary annulus. Catheter based interventions viz. repeat balloon dilatation or device closure of patent foramen ovale (PFO) may also be required in some patients.
Increased inter-equipment connectivity coupled with advances in Web technology allows ever escalating amounts of physiological data to be produced, far too much to be displayed adequately on a single computer screen. The consequence is that large quantities of insignificant data will be transmitted and reviewed. This carries an increased risk of overlooking vitally important transients. This paper describes a technique to provide an integrated solution based on a single algorithm for the efficient analysis, compression and remote display of long-term physiological signals with infrequent short duration, yet vital events, to effect a reduction in data transmission and display cluttering and to facilitate reliable data interpretation. The algorithm analyses data at the server end and flags significant events. It produces a compressed version of the signal at a lower resolution that can be satisfactorily viewed in a single screen width. This reduced set of data is initially transmitted together with a set of 'flags' indicating where significant events occur. Subsequent transmissions need only involve transmission of flagged data segments of interest at the required resolution. Efficient processing and code protection with decomposition alone is novel. The fixed transmission length method ensures clutter-less display, irrespective of the data length. The flagging of annotated events in arterial oxygen saturation, electroencephalogram and electrocardiogram illustrates the generic property of the algorithm. Data reduction of 87% to 99% and improved displays are demonstrated.
This paper attempts to improve the diagnostic quality of magnetic resonance (MR) images through application of lossy compression as a noise-reducing filter. The amount of imaging noise present in MR images is compared with the amount of noise introduced by the compression, with particular attention given to the situation where the compression noise is a fraction of the imaging noise. A popular wavelet-based algorithm with good performance, Set Partitioning in Hierarchical Trees (SPIHT), was employed for the lossy compression. Tests were conducted with a number of MR patient images and corresponding phantom images. Different plausible ratios between imaging noise and compression noise (ICR) were considered, and the achievable compression gain through the controlled lossy compression was evaluated. Preliminary results show that at certain ICR's, it becomes virtually impossible to distinguish between the original and compressed-decompressed image. Radiologists presented with a blind test, in certain cases, showed preference to the compressed image rather than the original uncompressed ones, indicating that under controlled circumstances, lossy image compression can be used to improve the diagnostic quality of the MR images.
This paper proposes a detection scheme for identifying stones in the biliary tract of the body, which is examined using magnetic resonance cholangiopancreatography (MRCP), a sequence of magnetic resonance imaging targeted at the pancreatobiliary region of the abdomen. The scheme enhances the raw 2D thick slab MRCP images and extracts the biliary structure in the images using a segment-based region-growing approach. Detection of stones is scoped within this extracted structure, by highlighting possible stones. A trained feedforward artificial neural network uses selected features of size and average segment intensity as its input to detect possible stones in MRCP images and eliminate false stone-like objects. The proposed scheme achieved satisfactory results in tests of clinical MRCP thick slab images, indicating potential for implementation in computer-aided diagnosis systems for the liver.
Tumors are generally difficult to detect in Magnetic Resonance (MR) images as they can be of varying intensities and do not appear as clear structures on these images. This difficulty is more prominent in MR Cholangiopancreatography (MRCP), which is the MR technology using a special sequence of T2-weighted imaging to identify the biliary tract, pancreatic duct, and gallbladder in the liver region, as MRCP images are more noisy in nature and are acquired for a more focused area with too much flexibility in position orientation for convenient computer-aided diagnosis. Based on the principle that the tumor mass manifests itself as blockage of the biliary tree structure, this paper introduces a technique that uses a region growing algorithm to identify discontinuities in the biliary tree as a means to preliminary detection of a possible tumor, in a fashion similar to the visual observation used by most radiologists in making their preliminary diagnosis. Through the use of appropriate image normalization, watershed segmentation, thresholding, rule-based region growing, and region analysis, the proposed technique is shown in this paper to be successful in identifying MRCP images with liver carcinoma from those with normal liver. Acquisition standardization, interactive image selection, and optimum image orientation will further enhance the accuracy of this proposed scheme for use in aiding clinical diagnosis at medical institutions.
In this paper, 3-D discrete Hartley transform is applied for the compression of two medical modalities, namely, magnetic resonance images and X-ray angiograms and the performance results are compared with those of 3-D discrete cosine and Fourier transforms using the parameters such as PSNR and bit rate. It is shown that the 3-D discrete Hartley transform is better than the other two transforms for magnetic resonance brain images whereas for the X-ray angiograms, the 3-D discrete cosine transform is found to be superior.
We describe a semi-automated technique for the quantitative assessment of breast density from digitized mammograms in comparison with patterns suggested by Tabar. It was developed using the MATLAB-based graphical user interface applications. It is based on an interactive thresholding method, after a short automated method that shows the fibroglandular tissue area, breast area and breast density each time new thresholds are placed on the image. The breast density is taken as a percentage of the fibroglandular tissue to the breast tissue areas. It was tested in four different ways, namely by examining: (i) correlation of the quantitative assessment results with subjective classification, (ii) classification performance using the quantitative assessment technique, (iii) interobserver agreement and (iv) intraobserver agreement. The results of the quantitative assessment correlated well (r2 = 0.92) with the subjective Tabar patterns classified by the radiologist (correctly classified 83% of digitized mammograms). The average kappa coefficient for the agreement between the readers was 0.63. This indicated moderate agreement between the three observers in classifying breast density using the quantitative assessment technique. The kappa coefficient of 0.75 for intraobserver agreement reflected good agreement between two sets of readings. The technique may be useful as a supplement to the radiologist's assessment in classifying mammograms into Tabar's pattern associated with breast cancer risk.
The effect-compartment concentration (C(e)) of a drug at a specific pharmacodynamic endpoint should be independent of the rate of drug injection. We used this assumption to derive an effect-compartment equilibrium rate constant (k(eo)) for propofol during induction of anesthesia, using a target controlled infusion device (Diprifusor). Eighteen unpremedicated patients were induced with a target blood propofol concentration of 5 microg x ml(-1) (group 1), while another 18 were induced with a target concentration of 6 microg x ml(-1) (group 2). The time at loss of the eyelash reflex was recorded. Computer simulation was used to derive the rate constant (k(eo)) that resulted in the mean C(e) at loss of the eyelash reflex in group 1 being equal to that in group 2. Using this population technique, we found the k(eo) to be 0.57 min(-1). The mean (SD) effect compartment concentration at loss of the eyelash reflex was 2.39 (0.70) microg x ml(-1). This means that to achieve a desired C(e) within 3 min of induction, the initial target blood concentration should be set at 1.67 times that of the desired C(e) for 1 min, after which it should revert to the desired concentration.