Experience-consistent modeling for radial basis function neural networks

We develop a new approach to the design of neural networks, which utilizes a collaborative framework of knowledge-driven experience. In contrast to the "standard" way of developing neural networks, which explicitly exploits experimental data, this approach incorporates a mechanism of knowledge-driven experience. The essence of the proposed scheme of learning is to take advantage of the parameters (connections) of neural networks built in the past for the same phenomenon (which might also exhibit some variability over time or space) for which are interested to construct the network on a basis of currently available data. We establish a conceptual and algorithmic framework to reconcile these two essential sources of information (data and knowledge) in the process of the development of the network. To make a presentation more focused and come up with a detailed quantification of the resulting architecture, we concentrate on the experience-based design of radial basis function neural networks (RBFNNs). We introduce several performance indexes to quantify an effect of utilization of the knowledge residing within the connections of the networks and establish an optimal level of their use. Experimental results are presented for low-dimensional synthetic data and selected datasets available at the Machine Learning Repository.     

Genome-enabled prediction of genetic values using radial basis function neural networks

The availability of high density panels of molecular markers has prompted the adoption of genomic selection (GS) methods in animal and plant breeding. In GS, parametric, semi-parametric and non-parametric regressions models are used for predicting quantitative traits. This article shows how to use neural networks with radial basis functions (RBFs) for prediction with dense molecular markers. We illustrate the use of the linear Bayesian LASSO regression model and of two non-linear regression models, reproducing kernel Hilbert spaces (RKHS) regression and radial basis function neural networks (RBFNN) on simulated data and real maize lines genotyped with 55,000 markers and evaluated for several trait-environment combinations. The empirical results of this study indicated that the three models showed similar overall prediction accuracy, with a slight and consistent superiority of RKHS and RBFNN over the additive Bayesian LASSO model. Results from the simulated data indicate that RKHS and RBFNN models captured epistatic effects; however, adding non-signal (redundant) predictors (interaction between markers) can adversely affect the predictive accuracy of the non-linear regression models.     

Identification of phytoplankton from flow cytometry data by using radial basis function neural networks.

We describe here the application of a type of artificial neural network, the Gaussian radial basis function (RBF) network, in the identification of a large number of phytoplankton strains from their 11-dimensional flow cytometric characteristics measured by the European Optical Plankton Analyser instrument. The effect of network parameters on optimization is examined. Optimized RBF networks recognized 34 species of marine and freshwater phytoplankton with 91. 5% success overall. The relative importance of each measured parameter in discriminating these data and the behavior of RBF networks in response to data from "novel" species (species not present in the training data) were analyzed.     

Training radial basis function neural networks: effects of training set size and imbalanced training sets

Obtaining training data for constructing artificial neural networks (ANNs) to identify microbiological taxa is not always easy. Often, only small data sets with different numbers of observations per taxon are available. Here, the effect of both size of the training data set and of an imbalanced number of training patterns for different taxa is investigated using radial basis function ANNs to identify up to 60 species of marine microalgae. The best networks trained to discriminate 20, 40 and 60 species respectively gave overall percentage correct identification of 92, 84 and 77%. From 100 to 200 patterns per species was sufficient in networks trained to discriminate 20, 40 or 60 species. For 40 and 60 species data sets an imbalance in the number of training patterns per species always affected training success, the greater the imbalance the greater the effect. However, this could be largely compensated for by adjusting the networks using a posteriori probabilities, estimated as network output values.     

Parameter estimation for stiff equations of biosystems using radial basis function networks

Background  

The modeling of dynamic systems requires estimating kinetic parameters from experimentally measured time-courses. Conventional global optimization methods used for parameter estimation, e.g. genetic algorithms (GA), consume enormous computational time because they require iterative numerical integrations for differential equations. When the target model is stiff, the computational time for reaching a solution increases further.     

Shape identification of electrocardiographic ST segment based on radial basis function neural network

The types of myocardial ischemia can be revealed by electrocardiographic (ECG) ST segment.Effective measurement and electrocardiographic analysis of ST as well as calculation of displacement and shape change of ST segment can help doctors diagnose coronary heart disease and myocardial ischemia,especially for asymptomatic myocardial ischemia.Therefore,it is a very important subject in clinical practice to measure and classify the ECG ST segment.In this paper,we introduce a computerized automatic identification method of the electrocardiographic ST segment shape with radial basis function neural network based on adaptive fuzzy system,which has a better effect than other methods.It helps to analyze the reason of the ST segment change and confirm the position of myocardial ischemia,and is useful for doctor diagnosis.     

Shape identification of electrocardiographic ST segment based on radial basis function neural network

The types of myocardial ischemia can be revealed by electrocardiographic (ECG) ST segment. Effective measurement and electrocardiographic analysis of ST as well as calculation of displacement and shape change of ST segment can help doctors diagnose coronary heart disease and myocardial ischemia, especially for asymptomatic myocardial ischemia. Therefore, it is a very important subject in clinical practice to measure and classify the ECG ST segment. In this paper, we introduce a computerized automatic identification method of the electrocardiographic ST segment shape with radial basis function neural network based on adaptive fuzzy system, which has a better effect than other methods. It helps to analyze the reason of the ST segment change and confirm the position of myocardial ischemia, and is useful for doctor diagnosis. Translated from Acta Biophysica Sinica, 2005, 21(6): 443–448 [译自: 生物物理学报]     

Adrenergic innervation of the lower urinary tract in cattle

Studies were carried out on 8 sexually immature male calves. Sections of the ureters, urinary bladder, and urethra were cut with a freezing microtome and the method of Ky?s?la et al. (1980) was used to visualize the adrenergic nerve fibres. It was found that bovine ureters possessed weak innervation; most of the nerves was located in the muscular membrane, and only in the paravesical part, sparse nerve fibres were found in the submucosa of this one. Apex of the urinary bladder was more weakly supplied with the adrenergic nerves than the corpus, whereas bladder's trigonum and cervix possessed numerous nerve fibres in both muscular and submucosal membranes. The distribution pattern of adrenergic nerves in the urethra was similar to that of urinary bladder's cervix. The presence of adregeneric nerve fibres was found in submucosal layer of both the urinary bladder and the urethra. Part of the nerves was connected with blood vessels of the organs under study.     

Minimally invasive procedures and medical management-their relative merits in treating lower urinary tract symptoms of benign prostatic hyperplasia

Evidence is reviewed supporting the safety and efficacy of minimally invasive transurethral microwave thermotherapy and medical management in patients with benign prostatic hyperplasia. Recent data indicate more pronounced long-term beneficial effects of microwave treatment. alpha-Blockade, however, offers more rapid onset of action than does microwave treatment. Neoadjuvant and adjuvant alpha-blocker therapy can accelerate symptom and flow rate improvement in patients undergoing microwave treatment. Compared with medical management, microwave treatment also appears to possess greater versatility, allowing patients who fall within a broad range of baseline symptom severities and prostate sizes to be treated with a high probability of success.     

Advances in stem cell therapy for the lower urinary tract

Lower urinary tract diseases are emotionally and financially burdensome to the individual and society. Current treatments are ineffective or symptomatic. Conversely, stem cells (SCs) are regenerative and may offer long-term solutions. Among the different types of SCs, bone marrow SCs (BMSCs) and skeletal muscle-derived SCs (SkMSCs) have received the most attention in pre-clinical and clinical trial studies concerning the lower urinary tract. In particular, clinical trials with SkMSCs for stress urinary incontinence have demonstrated impressive efficacy. However, both SkMSCs and BMSCs are difficult to obtain in quantity and therefore neither is optimal for the eventual implementation of SC therapy. On the other hand, adipose tissue-derived SCs (ADSCs) can be easily and abundantly obtained from "discarded" adipose tissue. Moreover, in several head-on comparison studies, ADSCs have demonstrated equal or superior therapeutic potential compared to BMSCs. Therefore, across several different medical disciplines, including urology, ADSC research is gaining wide attention. For the regeneration of bladder tissues, possible differentiation of ADSCs into bladder smooth muscle and epithelial cells has been demonstrated. For the treatment of bladder diseases, specifically hyperlipidemia and associated overactive bladder, ADSCs have also demonstrated efficacy. For the treatment of urethral sphincter dysfunction associated with birth trauma and hormonal deficiency, ADSC therapy was also beneficial. Finally, ADSCs were able to restore erectile function in various types of erectile dysfunction (ED), including those associated with diabetes, hyperlipidemia, and nerve injuries. Thus, ADSCs have demonstrated remarkable therapeutic potentials for the lower urinary tract.     

Distribution of serotonin-immunoreactive paraneurons in the lower urinary tract of dogs

Morphological and quantitative studies were made on serotonin-containing paraneurons throughout the lower urinary tract in male and female dogs. Using an anti-serotonin antiserum, the cells were consistently demonstrated to be dispersed in the epithelium from the vesico-urethral junction to the external urethral ostium. They occurred most frequently in the urethra proximal to the urogenital diaphragm in both sexes. The total number of the serotonin-immunoreactive cells in the urethra was estimated to be 36.2 X 10(4) (SD 9.9 X 10(4] in the male (n = 3) and 15.6 X 10(4) (SD 2.1 X 10(4] in the female (n = 3). Besides the urethra, the prostate and vaginal vestibule contained several serotonin-immunoreactive cells. The urethral serotonin cells were basically bipolar basal-granulated cells that extended the basal cytoplasm to the basement membrane and reached the lumen with an apical process. Modified cell shapes were, however, also frequent, and included bifurcated apical and/or basal processes or a laterally directed basal process. Occasional serotonin cells possessed a threadlike basal process with varicosities and a terminal bouton, reminiscent of a neuronal process. Immunoreactivity for chromogranin A, a carrier protein common to endocrine paraneurons, was demonstrated in all of the urethral serotonin cells. The chromogranin A-immunoreactive granules accumulated more densely in the basal and perinuclear regions of the cell. It is hypothesized that the serotonin-immunopositive paraneurons may receive chemical and/or physical information from urine and, in response to it, secrete serotonin which presumably causes the contraction of the musculature of the urethra.     

A computer model for describing the effect of urethral afferents on simulated lower urinary tract function

A computer model of mechanical properties of the bladder, the urethra and the rhabdosphincter, as well as their neural control is presented in this paper. The model has a rather simple design and processes sensory information from both the bladder wall tension and urethral stretch. It is assumed that afferent signals from the urethra are involved in a sacral excitatory reflex and a supraspinal inhibitory reflex. Pressure and flow signals that resemble experimentally measured normal human behaviour could be simulated with this model. From these simulations the relation between the neural control mechanisms used in the model and the neural control mechanism in vivo cannot be judged entirely because similar behaviour could be simulated with models that are bas ed on different neural control mechanisms. Also behaviour that resembles detrusor overactivity was simulated with our model after an externally induced rise in detrusor pressure was added. Detrusor overactivity, sometimes in combination with urethral relaxation, can occur during a urodynamic investigation. A possible explanation for this detrusor overactivity might be that the micturition reflex is triggered by unknown disturbances and is inhibited immediately after by the same mechanism that normally ceases voiding. The described model provides such a mechanism. Based on these simulations, therefore, it is concluded that urethral afferent signals might be important in lower urinary tract control.     

Modelling crystal aggregation and deposition in the catheterised lower urinary tract

Urethral catheters often become encrusted with crystals of magnesium struvite and calcium phosphate. The encrustation can block the catheter, which can cause urine retention in the bladder and reflux into the kidneys. We develop a mathematical model to investigate crystal deposition on the catheter surface, modelling the bladder as a reservoir of fluid and the urethral catheter as a rigid channel. At a constant rate, fluid containing crystal particles of unit size enters the reservoir, and flows from the reservoir through the channel and out of the system. The crystal particles aggregate, which we model using Becker–Döring coagulation theory, and are advected through the channel, where they continue to aggregate and are deposited on the channel’s walls. Inhibitor particles also enter the reservoir, and can bind to the crystals, preventing further aggregation and deposition. The crystal concentrations are spatially homogeneous in the reservoir, whereas the channel concentrations vary spatially as a result of advection, diffusion and deposition. We investigate the effect of inhibitor particles on the amount of deposition. For all parameter values, we find that crystals deposit along the full length of the channel, with maximum deposition close to the channel’s entrance.     

Psychological aspects of lower urinary tract infections in women.

OBJECTIVE--To determine whether women with the urethral syndrome can be distinguished from those with urinary tract infection by case notes, clinical symptoms, or psychiatric state. DESIGN--Longitudinal survey of consecutive women presenting with dysuria and frequency. SETTING--General practice and community. SUBJECTS--58 patients with the urethral syndrome and 44 patients with a urinary tract infection, mean age 39.9 years. MAIN OUTCOME MEASURES--Results of analysis of serial midstream urine specimens, patients'' self rated physical symptoms and responses to 60 item general health questionnaire at presentation and after resolution of symptoms, and results of psychiatric assessment with the clinical psychiatric interview. RESULTS--4 of 42 patients with a urinary tract infection had recently changed sexual partner compared with none of 58 with the urethral syndrome. Dysuria and nocturia were more common in patients with urinary tract infections than those with the urethral syndrome (mean (SD) score for dysuria 5.37 (2.39) v 4.57 (2.13), p less than 0.05; nocturia in 39/44 (88%) patients v 40/58 (69%), chi 2 = 5.5, p less than 0.02). Both groups showed transient high levels of distress which resolved with the physical symptoms, but no psychiatric difference distinguished them. CONCLUSION--The urethral syndrome is not associated with increased psychiatric morbidity.     

Radial basis function neural networks for modeling growth rates of the basidiomycetes Physisporinus vitreus and Neolentinus lepideus

A radial basis function (RBF) neural network was developed and compared against a quadratic response surface (RS) model for predicting the specific growth rates of the biotechnologically important basidiomycetous fungi, Physisporinus vitreus and Neolentinus lepideus, under three environmental conditions: temperature (10–30 °C), water activity (0.950–9.998), and pH (4–6). Both the RBF network and polynomial RS model were mathematically evaluated against experimental data using graphical plots and several statistical indices. The evaluation showed that both models gave reasonably good predictions, but the performance of the RBF neural network was superior to that of the classical statistical method for all three data sets used (training, testing, full). Sensitivity analysis revealed that of the three experimental factors the most influential on the growth rate of P. vitreus was water activity, followed by temperature and pH to a lesser extent. In contrast, temperature in particular and then water activity were the key determinants of the development of N. lepideus. RBF neural networks could be a powerful technique for modeling fungal growth behavior under certain parameters and an alternative to time-consuming, traditional microbiological techniques.