Neural networks with dynamic synapses

Transmission across neocortical synapses depends on the frequency of presynaptic activity (Thomson & Deuchars, 1994). Interpyramidal synapses in layer V exhibit fast depression of synaptic transmission, while other types of synapses exhibit facilitation of transmission. To study the role of dynamic synapses in network computation, we propose a unified phenomenological model that allows computation of the postsynaptic current generated by both types of synapses when driven by an arbitrary pattern of action potential (AP) activity in a presynaptic population. Using this formalism, we analyze different regimes of synaptic transmission and demonstrate that dynamic synapses transmit different aspects of the presynaptic activity depending on the average presynaptic frequency. The model also allows for derivation of mean-field equations, which govern the activity of large, interconnected networks. We show that the dynamics of synaptic transmission results in complex sets of regular and irregular regimes of network activity.     

Neural networks as models of psychopathology

Neural network modeling is situated between neurobiology, cognitive science, and neuropsychology. The structural and functional resemblance with biological computation has made artificial neural networks (ANN) useful for exploring the relationship between neurobiology and computational performance, i.e., cognition and behavior. This review provides an introduction to the theory of ANN and how they have linked theories from neurobiology and psychopathology in schizophrenia, affective disorders, and dementia.     

Neural networks as a prognostic tool for patients with non-small cell carcinoma of the lung

Patients with non-small cell carcinoma of the lung (NSCLC) have a poor prognosis (64 and 41% survival rates in Stages I and II). It is currently not possible to predict which patients with Stage I or II NSCLC will survive the disease. Sixty-seven patients with NSCLC, including 49 patients with Stage I NSCLC and 18 with Stage II disease (11 with squamous cell carcinomas, 35 with adenocarcinomas, and 21 with large cell carcinomas) were treated with lobectomy and followed for a minimum of 5 years. The tumors were studied with DNA flow cytometry and quantitative immunocytochemical studies for proliferation cell nuclear antigen, p53 protein, and MIB-1. The data were analyzed with backpropagation neural networks, univariate analysis of variance, the Kaplan-Meier survival method, and Cox proportional hazards model. The dependent variables were "free of disease" and "recurrence or dead from disease." Twenty neural network models were trained, using all cases but one, after 1883 to 2000 training cycles. At 5 years, 30 patients were free of disease and 37 were dead or had recurrence. Proliferating cell nuclear antigen was the only statistically significant prognostic factor by univariate analysis of variance and Cox proportional hazards analysis. The S phase was statistically significant by univariate analysis of variance (P <.05). All of the 20 models classified the test cases correctly. Study with backpropagation neural networks using multiple prognostic features from patients with NSCLC suggests that this technology might be useful for prediction of survival. This preliminary study must be validated with data from a larger group of patients with NSCLC before its clinical adequacy is established.     

基于FPGA的神经网络硬件实现方法

提出了一种可以灵活适应不同的工程应用中神经网络在规模、拓扑结构、传递函数和学习算法上的变化,并能及时根据市场需求快速建立原型的神经网络硬件可重构实现方法.对神经网络的可重构特征进行了分析,提出了三种主要的可重构单元;研究了可重构的脉动体系结构及BP网络到该结构映射算法;探讨了具体实现的相关问题.结果表明,这种方法不仅灵活性强,其实现的硬件也有较高的性价比,使用一片FPGA中的22个乘法器工作于100MHz时,学习速度可达432MCUPS.     

Neural networks for recognizing patterns in cardiotocograms

Leptomeningeal metastases (LM) is a common problem in neuro-oncology occurring in approximately 5% of all patients with cancer. Notwithstanding frequent focal signs and symptoms in LM, LM is a disease affecting the entire neuraxis and therefore staging and treatment need encompass all cerebrospinal fluid (CSF) compartments. Central nervous system (CNS) staging of LM includes contrast enhanced cranial computerized tomography (CE-CT) or magnetic resonance imaging (MR-Gd), contrast enhanced spine magnetic resonance imaging (MR-S) or computerized tomographic myelography (CT-M) and radionuclide CSF flow study (FS). Treatment of LM involves involved-field radiotherapy of bulky or symptomatic disease sites and intra-CSF drug therapy. The inclusion of concomitant systemic therapy may benefit patients with LM and may obviate the need for intra-CSF chemotherapy. At present, intra CSF drug therapy is confined to three chemotherapeutic agents (i.e. methotrexate, cytosine arabinoside and thio-TEPA) administered by a variety of schedules either by intralumbar or intraventricular drug delivery. Although treatment of LM is palliative with an expected median patient survival of 6 months, it often affords stabilization and protection from further neurologic deterioration in patients with LM.     

Neural networks: a bridge between neuroscience and psychology

Between the anatomy and physiology of neurons on the one hand, and the study of perception and personality on the other, there is a vast gap in our understanding. Although, today, most neuroscientists say it is axiomatic that the phenomena of mind result from the operation of the brain, they could not tell you how.     

Neural control of cochlear blood flow

Effects of intraarterial and intravenous injections of autonomic nervous system agents on cochlear blood flow were studied in order to investigate the neural control of the inner ear vessels. Blood flow changes in the inner ear of the guinea pig were measured with an electrical impedance plethysmograph. Rather weak control of the vertebrobasilar and labyrinthine arteries by the sympathetic nervous system of the alpha-receptor type did appear to exist. Beta-receptors of the sympathetic nerve appeared to be non-existent in the cochlear vessels, and parasympathetic modulation was not evident.     

Neural networks for the analysis of small pulmonary nodules

PURPOSE: Small pulmonary nodules can be readily detected by computed tomography (CT). The goal of this detection is to diagnose early lung cancer as the five year survival at this early stage is over 70% in contradistinction to the overall 5-year survival of around 10%. Critical to the efficacy of CT for early lung cancer detection is the ability to distinguish between benign and malignant nodules. We explored the usefulness of neural networks (NNs) to help in this differentiation. METHODS: CT images of 28 pulmonary nodules, 14 benign and 14 malignant, each having a diameter less than 3 cm were selected. All were sufficiently malignant in appearance to require needle biopsy and surgery. The statistical-multiple object detection and location system (S-MODALS) NN technique developed for automatic target recognition (ATR) was used to differentiate between these benign and malignant nodules. RESULTS: S-MODALS was able to correctly identify all but three benign nodules. S-MODALS classified a nodule as malignant because it looked similar to other malignant nodules. It identified the most similar nodules to display them to the radiologist. The specific features of the nodule that determined its classification were also shown, so that S-MODALS is not simply a "black box" technique but gives insight into the NN diagnostics. CONCLUSION: This initial evaluation of S-MODALS NNs using pulmonary nodules whose CT features were very suspicious for lung cancer demonstrated the potential to reduce the number of biopsies without missing malignant nodules. S-MODALS performed well, but additional optimization of the techniques specifically for CT images would further enhance its performance.     

Neural networks in the analysis of episodic growth hormone release

Pulsatile secretion of growth hormone (GH) has been observed in healthy controls as well as acromegalic patients. In healthy adults, highly regulated secretory pulses of GH occur 4-8 times within 24 h. This episodic pattern of secretion seems to be related to the optimal induction of physiological effects at the peripheral level. In contrast to normal subjects, acromegalic patients demonstrate an irregular pattern of excessive GH release. This pattern of secretion is responsible for many systemic effects, such as the stimulation of connective tissue growth, cardiovascular and cerebrovascular disease, diabetes mellitus and arthritis. Standard methods for the analysis of pulsatile patterns of hormone secretion did not consistently separate the temporal dynamics of GH release in healthy controls and acromegalic patients under various study conditions. Using the cutting edge technology of artificial neural networks for time series prediction, we were able to achieve significant separation of both groups under various conditions by means of the predictability of their GH secretory dynamics. Improving the predictive results by using a more refined system of multiple neural networks acting in parallel (adaptive mixtures of local experts), we found that this system performed a self-organized segmentation of hormone pulsatility. It separated phases of secretory bursts and quiescence without any prior knowledge of the form of a GH pulse or a model of secretion. Comparing the predictive results for the GH dynamics with those for computer-stimulated stochastic processes, we were able to define the irregular pattern of GH secretion in acromegaly as a random autonomous process. The introduction of neural networks to the analysis of dynamic endocrine systems might help to expand the existing analytical approaches beyond counting frequency and amplitude of hormone pulses.     

Neural cell classification by wavelets and multiscale curvature

It has been postulated that the aberrant projection of sympathetic axons to individual primary sensory neurons may provide the morphological basis for pain-related behaviors in rat models of chronic pain syndrome. Since nerve growth factor (NGF) can elicit the collateral sprouting of noradrenergic sympathetic terminals, it might be predicted that NGF plays a role in mediating the sprouting of sympathetic axons into sensory ganglia. Using a line of transgenic mice overexpressing NGF among glial cells, it was first found that trigeminal ganglia from adult transgenic mice possessed significantly higher levels of NGF protein in comparison to age-matched wild-type mice; as well, detectable levels of NGF mRNA transgene expression were present in both the ganglia and brain stem. Within the trigeminal ganglia, a small proportion of the sensory neuronal population stained immunohistochemically for NGF; a higher percentage of NGF-positive neurons was evident in transgenic mice. New sympathetic axons extended into the trigeminal ganglia of transgenic mice only and formed perineuronal plexuses surrounding only those neurons immunostained for NGF. In addition, such plexuses were accompanied by glial processes from nonmyelinating Schwann cells. From these data, we propose that accumulation of glial-derived NGF by adult sensory neurons and its putative release into the ganglionic environment induce the directional growth of sympathetic axons to the source of NGF, namely, the cell bodies of primary sensory neurons.