SEARCH · PubMed Health

Results for “Multilayer Perceptrons”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

304 records · Page 17Linked to original sources

PubMed1993

A computerized neural network method for pattern recognition of cocaine signatures.

This article describes a practical procedure for rapidly searching a large database of cocaine signatures to identify database entries that closely resemble a given reference cocaine exhibit using a personal computer (PC). The procedure takes advantage of the pattern recognition capability of the multilayer perceptron neural network to identify similar cocaine signatures. A PC-based software implementation is now being used on a daily basis at the North Carolina State Bureau of Investigation (NCSBI) to aid forensic experts in identifying signatures that originate from the same batch. Intelligence reports generated from database searches have been useful to undercover agents in the field who are striving to build drug related conspiracy cases. This software was developed as a collaborative effort between the NCSBI and the Center for Systems and Engineering of the Research Triangle Institute.

Chromatography, Gas↗

PubMed1995

Improved classification of psychiatric mood disorders using a feedforward neural network.

Our aim was to use a neural network model for classifying psychiatric mood disorders. In approaching this interdisciplinary task, the concern was not to build a more biologically plausible model, but to test the computational capabilities of an existent one. The multilayer perceptron was chosen as a model, trained and tested using both theoretical and real cases. The classification proved to be good, but we are aware of the fact that both training and testing should be performed in the future only on real data, as the way to get a true picture of the system's performance.

Adult↗

PubMed1995

Application of optimized pattern recognition units in EEG analysis: common optimization of preprocessing and weights of neural networks as well as structure optimization.

The main goal of this study is to demonstrate the possibility of training the Neural Network (multilayer perceptron) classifier and preprocessing units simultaneously, i.e., that properties of preprocessing are chosen automatically during the training phase. In the first realization step, adaptive recursive estimation of the power within a frequency band was used as a preprocessing unit. To improve the efficiency of special units, the power and momentary frequency estimation was replaced by methods that are based on adaptive Hilbert transformers. The strategy was developed to obtain optimized recognition units that can be efficiently integrated into strategies for monitoring the cerebral status of neonates. Therefore, applications (e.g., in neonatal EEG pattern recognition) will be shown. Additionally, a method of minimizing the error function was used, where this minimization is based on optimizing the network structure. The results of structure optimization in the field of EEG pattern recognition in epileptic patients can be demonstrated.

Algorithms↗

PubMed1995

ISDN based teleradiology and image analysis with the software system KAMEDIN.

This contribution describes the software system KAMEDIN (Kooperatives Arbeiten und MEdizinische Diagnostik auf Innovativen Netzen) that is designed as an ISDN based computer supported cooperative work (CSCW) tool for usage in medical diagnostics. Medical image data from various sources (for example CT and MR) can be interchanged and analyzed in bilateral teleconferences via ISDN. During a cooperative session, user interactions for image processing etc., are synchronized and performed on both workstations, Features like telepointing, remote control, and audio connection enhance communication quality. With ISDN as a transmission line, widespread availability and low communication costs are achieved. Further, automatic tissue labeling in intracranial MR data can be invoked. For this purpose, artificial neural network classifiers such as multilayer perceptron and Kohenen feature map are integrated. Classification results can be viewed as 3D-reconstructions.

Brain Diseases↗

PubMed1997

Automated interpretation of myocardial SPECT perfusion images using artificial neural networks.

UNLABELLED: The purpose of this study was to develop a computer-based method for automatic detection and localization of coronary artery disease (CAD) in myocardial bull's-eye scintigrams. METHODS: A population of 135 patients who had undergone both myocardial 99mTc-sestamibi rest-stress scintigraphy and coronary angiography within 3 mo was studied. Different image data reduction methods, including pixel averaging and two-dimensional Fourier transform, were applied to the bull's-eye scintigrams. After a quantitative and qualitative evaluation of these methods, 30 Fourier components were chosen as inputs to multilayer perceptron artificial neural networks. The networks were trained to detect CAD in two vascular territories, using coronary angiography as gold standard. A "leave one out" procedure was used for training and evaluation. The performance of the networks was compared to those of two human experts. RESULTS: One of the human experts detected CAD in one of two vascular territories, with a sensitivity of 54.4% at a specificity of 70.5%. The sensitivity of the networks was significantly higher at that level of specificity (77.2%, p = 0.0022). The other expert had a sensitivity of 63.2% at a specificity of 61.5%. The networks had a sensitivity of 77.2% (p = 0.038) at this specificity level as well. The differences in sensitivity between human experts and networks for the other vascular territory were all less than 6% and were not statistically significant. CONCLUSION: Artificial neural networks can detect CAD in myocardial bull's-eye scintigrams with such a high accuracy that the application of neural networks as clinical decision support tools appears to have significant potential.

Coronary Angiography↗

PubMed1998

Feature-extraction from endopeptidase cleavage sites in mitochondrial targeting peptides.

Cleavage sites in nuclear-encoded mitochondrial protein targeting peptides (mTPs) from mammals, yeast, and plants have been analysed for characteristic physicochemical features using statistical methods, perceptrons, multilayer neural networks, and self-organizing feature maps. Three different sequence motifs were found, revealing loosely defined arginine motifs with Arg in positions -10, -3, and -2. A self-organizing feature map was able to cluster these three types of endopeptidase target sites but did not identify any species-specific characteristics in mTPs. Neural networks were used to define local sequence features around precursor cleavage sites.

Amino Acid Sequence↗

PubMed1998

The recursive deterministic perceptron neural network.

We introduce a feedforward multilayer neural network which is a generalization of the single layer perceptron topology (SLPT), called recursive deterministic perceptron (RDP). This new model is capable of solving any two-class classification problem, as opposed to the single layer perceptron which can only solve classification problems dealing with linearly separable sets (two subsets X and Y of R(d) are said to be linearly separable if there exists a hyperplane such that the elements of X and Y lie on the two opposite sides of R(d) delimited by this hyperplane). We propose several growing methods for constructing a RDP. These growing methods build a RDP by successively adding intermediate neurons (IN) to the topology (an IN corresponds to a SLPT). Thus, as a result, we obtain a multilayer perceptron topology, which together with the weights, are determined automatically by the constructing algorithms. Each IN augments the affine dimension of the set of input vectors. This augmentation is done by adding the output of each of these INs, as a new component, to every input vector. The construction of a new IN is made by selecting a subset from the set of augmented input vectors which is LS from the rest of this set. This process ends with LS classes in almost n-1 steps where n is the number of input vectors. For this construction, if we assume that the selected LS subsets are of maximum cardinality, the problem is proven to be NP-complete. We also introduce a generalization of the RDP model for classification of m classes (m>2) allowing to always separate m classes. This generalization is based on a new notion of linear separability for m classes, and it follows naturally from the RDP. This new model can be used to compute functions with a finite domain, and thus, to approximate continuous functions. We have also compared - over several classification problems - the percentage of test data correctly classified, or the topology of the 2 and m classes RDPs with that of the backpropagation (BP), cascade correlation (CC), and two other growing methods.

Journal Article↗

PubMed1996

Application of artificial neural networks to clinical pharmacology.

Drug dosages and drug choices are determined by a knowledge of the drug's pharmacokinetics and pharmacodynamics. Often, insufficient information is available to determine the pharmacokinetics of a drug or which drug will have a desired effect for an individual patient. We propose that a form of nonlinear regression, an artificial neural network, can be used. We have demonstrated this use with 2 examples. In the first example we use a neural network to predict gentamicin peak and trough concentrations from routine therapeutic drug monitoring. In the second example we predict delayed renal allograft function as a guide for induction of immunosuppression therapy. Predictions were made using a multilayer feedforward perceptron and compared to nonlinear mixed effect modeling (NONMEM) and logistic regression. Neural network peak and trough gentamicin predictions were more precise and less biased than control predictions made using NONMEM. Prediction error for peak serum concentrations averaged 16.5% for the neural networks and 18.6% for NONMEM. Prediction error for trough concentrations were 48.3% for neural networks and 59.0% for NONMEM. When used for the prediction of delayed graft function, the neural network correctly predicted immediate graft function 73% of the time and delayed graft function 65% of the time. For those patients predicted to develop delayed graft function, alternate induction using anti-lymphocyte globulin may be indicated. These 2 examples demonstrate how an artificial neural network may be applied to predictions in clinical pharmacology.

Anti-Bacterial Agents↗

PubMed2000

Acoustic identification of twelve species of echolocating bat by discriminant function analysis and artificial neural networks.

We recorded echolocation calls from 14 sympatric species of bat in Britain. Once digitised, one temporal and four spectral features were measured from each call. The frequency-time course of each call was approximated by fitting eight mathematical functions, and the goodness of fit, represented by the mean-squared error, was calculated. Measurements were taken using an automated process that extracted a single call from background noise and measured all variables without intervention. Two species of Rhinolophus were easily identified from call duration and spectral measurements. For the remaining 12 species, discriminant function analysis and multilayer back-propagation perceptrons were used to classify calls to species level. Analyses were carried out with and without the inclusion of curve-fitting data to evaluate its usefulness in distinguishing among species. Discriminant function analysis achieved an overall correct classification rate of 79% with curve-fitting data included, while an artificial neural network achieved 87%. The removal of curve-fitting data improved the performance of the discriminant function analysis by 2 %, while the performance of a perceptron decreased by 2 %. However, an increase in correct identification rates when curve-fitting information was included was not found for all species. The use of a hierarchical classification system, whereby calls were first classified to genus level and then to species level, had little effect on correct classification rates by discriminant function analysis but did improve rates achieved by perceptrons. This is the first published study to use artificial neural networks to classify the echolocation calls of bats to species level. Our findings are discussed in terms of recent advances in recording and analysis technologies, and are related to factors causing convergence and divergence of echolocation call design in bats.

Acoustics↗

PubMed1997

Removing irrelevant features in neural network classification using evolutionary computations.

Evolutionary artificial neural networks (EANN) are a new paradigm that refers to a special class of artificial neural networks (ANN) in which evolution is another fundamental form of adaptation in addition to learning. Evolution can be introduced at various levels of ANN. It can be used to evolve weights, architectures and learning parameters. Evolutionary computations are population-based search methods that have shown promise in many similarly complex tasks. This paper presents an application of evolutionary programming for simultaneously inducing the input structure and weights evolving for multilayer feed-forward perceptrons (MLP) with standard sigmoidal activation function.

Algorithms↗

PubMed2004

Comparison of machine learning techniques with classical statistical models in predicting health outcomes.

Several machine learning techniques (multilayer and single layer perceptron, logistic regression, least square linear separation and support vector machines) are applied to calculate the risk of death from two biomedical data sets, one from patient care records, and another from a population survey. Each dataset contained multiple sources of information: history of related symptoms and other illnesses, physical examination findings, laboratory tests, medications (patient records dataset), health attitudes, and disabilities in activities of daily living (survey dataset). Each technique showed very good mortality prediction in the acute patients data sample (AUC up to 0.89) and fair prediction accuracy for six year mortality (AUC from 0.70 to 0.76) in individuals from epidemiological database surveys. The results suggest that the nature of data is of primary importance rather than the learning technique. However, the consistently superior performance of the artificial neural network (multi-layer perceptron) indicates that nonlinear relationships (which cannot be discerned by linear separation techniques) can provide additional improvement in correctly predicting health outcomes.

Aged↗

PubMed1998

The fractional correction rule: a new perspective.

In this paper we discover and explore a useful property of the fractional correction rule, a variation of the perceptron learning rule, for a single neural unit. We rename this rule the projection learning rule (PrLR), which seems more appropriate because of the technique that we use to prove convergence. We state and prove a more powerful convergence theorem and establish the link with S. Agmon's work (1954) on linear inequalities. The hallmark of this rule is that if the problem is not linearly separable then the proposed rule will always converge to the origin of the weight-space. This points out that appropriate nonlinear methods (e.g., multilayer neural network, nonlinear transformation on input-space etc.) should be used to address this problem. On the other hand, if the patterns are linearly separable, the performance of this rule is equivalent to that of the perceptron rule. A theoretical investigation of this rule leads to interesting observations. We present experimental results with linearly separable as well as linearly non-separable data using the PrLR and compare its performance with that of the perceptron rule.

Journal Article↗

PubMed2001

Computerized classification of suspicious regions in chest radiographs using subregion Hotelling observers.

We propose to investigate the use of subregion Hotelling observers (SRHOs) in conjunction with perceptrons for the computerized classification of suspicious regions in chest radiographs for being nodules requiring follow up. Previously, 239 regions of interest (ROIs), each containing a suspicious lesion with proven classification, were collected. We chose to investigate the use of SRHOs as part of a multilayer classifier to determine the presence of a nodule. Each SRHO incorporates information about signal, background, and noise correlation for classification. For this study, 225 separate Hotelling observers were set up in a grid across each ROI. Each separate observer discriminates an 8 by 8 pixel area. A round robin sampling scheme was used to generate the 225 features, where each feature is the output of the individual observers. These features were then rank ordered by the magnitude of the weights of a perceptron. Once rank ordered, subsets of increasing number of features were selected to be used in another perceptron. This perceptron was trained to minimize mean squared error and the output was a continuous variable representing the likelihood of the region being a nodule. Performance was evaluated by receiver operating characteristic (ROC) analysis and reported as the area under the curve (Az). The classifier was optimized by adding additional features until the Az declined. The optimized subset of observers then were combined using a third perceptron. A subset of 80 features was selected which gave an Az of 0.972. Additionally, at 98.6% sensitivity, the classifier had a specificity of 71.3% and increased the positive predictive value from 60.7% to 84.1 %. Preliminary results suggest that using SRHOs in combination with perceptrons can provide a successful classification scheme for pulmonary nodules. This approach could be incorporated into a larger computer aided detection system for decreasing false positives.

Algorithms↗

PubMed1999

Color image correction for scanner and printer using B-spline CMAC neural networks.

The process of eliminating the color errors from the gamut mismatch, resolution conversion, and nonlinearity between scanner and printer is usually recognized as an essential issue of color reproduction. This paper presents a new formulation based on the generalized inverse plant control for the color error reduction process. In our formulation, the printer input and scanner output correspond to the input and output of a system plant, respectively. Obviously, if the printer input equals the scanner output, then there are no color errors involved in the entire system. In other words, the plant becomes an identity system. To achieve this goal, a plant generalized inverse should be identified and added to the original system. Since the system of a combination of both scanner and printer is highly nonlinear, CMAC-based neural networks, which have the capability to learn arbitrary nonlinearity, are applied to identify the plant generalized inverse. CMAC network is a perceptron-like feedforward structure with associative memory properties. Its memory requirements can be greatly reduced by the use of hash coding techniques. In order for CMAC networks to construct high-order, smooth, nonlinear plant inverse, more general CMAC addressing schemes have been proposed in conjunction with use of B-spline receptive functions. It is shown that B-spline CMAC networks learn orders of magnitude more rapidly than typical implementations of back propagation in the multilayered neural networks, due to the local nature of its weighting updating and the finite support of B-spline receptive field functions. Finally, a number of test samples are conducted to verify the effectiveness of the proposed method.

Color↗

PubMed1978

Learning processes in multilayer threshold nets.

An algorithm of learning in multilayer threshold nets without feedbacks is proposed. The net is built of threshold elements with binary inputs. During a learning process each input vector chi is accompanied by a teacher's decision omega (omega epsilon(1,...,M)). The pairs (chi[n], omega[n]) appear in successive steps independently according to some unknown stationary distribution p(chi, omega). The problem of learning of a threshold net has been decomposed to a series of problems of learning of the threshold elements. The proposed learning algorithm of the threshold elements has a perceptron-like form. It was proven that a decision rule of the threshold net stabilizes after a finite number of steps. For definite classes (p(chi,omega))K of distributions p(chi, omega), an optimal decision rule stabilizes after a finite number of steps. These classes (p(chi, omega))K also contain distributions describing learning processes with perturbations.

Feedback↗

PubMed2003

[Use of multilayer perception artificial neutral networks for the prediction of the probability of malignancy in adnexal tumors].

BACKGROUND: Advanced statistical methods are currently are more often used in the prediction of ovarian malignancy when adnexal tumor is detected. These methods include logistic regression analysis and artificial neural networks (ANN's), i.e. computer programs which are capable of learning from presented data and further predict various events, such as clinical diagnosis or outcome of a given treatment. MATERIALS AND METHODS: We have analyzed data of 307 women with adnexal tumors who were operated in the Ist Dept. of Gynecology, Medical University in Lublin between 2000-2002. Following clinical and sonographic variables were included: age, menopausal status, serum CA-125, bilaterality, tumor size and volume, papillary projections, septa, solid parts presence, Doppler blood flow indices (PI, RI, Vmax), and subjective-color Doppler score. A multiple layer perceptron (MLP) neural network with 13 input variables, 11 hidden neurons and one output variable was constructed to assess probability of malignancy in each women (Statistica v. 6.0 for Windows, Statsoft, USA). Sensitivity, specificity and accuracy of the model were calculated. Receiver-Operating Characteristics curves were generated and corresponding Areas Under ROC Curves (AUROC's) for all diagnostic tests were compared. RESULTS: Final histologic examination revealed 228 (74.3%) benign tumors and 79 (25.7%) malignant masses including 21 women with FIGO stage I ovarian cancer. With a 75% cut-off probability of malignancy level the sensitivity and specificity of the best network in the testing set was 96.7% and 100%, respectively. In the validation set the corresponding values of sensitivity and specificity were 82.3% and 97.5%. The highest of all used tests AUROC equal to 0.9749 was found for the ANN predictive model. CONCLUSIONS: ANN may help in the extraction of the most useful predictive clinical and ultrasound data. The sensitivity and specificity of the ANN's generated model were higher than currently used single clinical and diagnostic tests. However, a prospective testing in a new, much larger group of women with adnexal tumors is essential for the clinical usefulness of the proposed statistical model.

Adult↗