Synaptic changes in the brain of subjects with schizophrenia

Clinical, epidemiological, neuroimaging and postmortem data all suggest schizophrenia is a neurodevelopmental disorder, and that synaptic disturbances might play a critical role in developing the disease. In 1982, Feinberg proposed that the schizophrenia might arise as a result of abnormal synaptic pruning. His hypothesis has survived 40 years of accumulated data, and we review the critical findings related to synaptic dysfunction of schizophrenia. While it is clear that synaptic disturbances are integral and important for understanding the pathophysiology of schizophrenia, it has also become obvious that synaptic disturbances cannot be studied and understood as an independent disease hallmark, but only as a part of a complex network of homeostatic events. Development, glial-neural interaction, changes in energy homeostasis, diverse genetic predisposition, neuroimmune processes and environmental influences all can tip the delicate homeostatic balance of the synaptic morphology and connectivity in a uniquely individual fashion, thus contributing to the emergence of the various symptoms of this devastating disorder. Finally, we argue that based on a predominant change in gene expression pattern we can broadly sub-stratify schizophrenia into “synaptic” “oligodendroglial”, “metabolic” and “inflammatory” subclasses.     

基于限制模型规模和声学置信度的关键词检出方法

在基于令牌传递算法的关键词检出技术中,为改进实时性,本文首先从限制模型规模的角度,提出了限制上下文相关的词内相关音素模型.针对误识率高的问题,提出了基于声学置信度的关键词确认方法,并实现了多次解码机制,提高了识别性能.其次,从改进解码算法的角度,研究了剪枝和控制最大激活模型数两种策略对识别性能的影响,并结合确认机制进行关键词检出,获得了满意的结果.     

基于人工神经网络的荧光光谱识别

目的利用人工神经网络的方法对杂质气体的荧光光谱进行识别。方法选用三层前馈型单输出的神经网络结构,学习算法选用自适应的RLS（Recursive Least Square）算法,在对基于RLS算法的三层前馈单输出的神经网络学习训练完成后,引入互联权重删剪算法对网络的输入层进行删剪。网络的结构不但得到了优化,而且计算复杂度也大大降低。提高网络的泛化能力;最后利用优化后的网络对测试样本进行识别。结果仿真实验表明,与删剪前的网络结构相比,在降低了网络的计算复杂度的同时,删剪优化后的正确识别率能够达到100%。结论利用RLS算法能够提高网络的运算速度,在此基础上的删剪算法能够降低计算的复杂度,并且能够对冗余信息进行了剔除,进而提高整个网络的识别能力。     

施肥、修剪等措施对凹叶厚朴中厚朴酚、和厚朴酚含量累积的影响

目的：提高凹叶厚朴的品质,缩短生产周期,为凹叶厚朴种植提供科学参考依据。方法：本次实验通过采用施肥、修枝、环割、铁丝环扎对凹叶厚朴进行处理,采用HPLC测定两年后厚朴中厚朴酚、和厚朴酚的含量,计算总酚含量增长率。结果：凹叶厚朴中总酚含量累积增长（1,）与肥料中的氮肥、磷肥正相关,与钾肥负相关,回归方程为：Y=1．804＋2．470N＋2．746P-15．616K（r=0．982）。修枝处理的凹叶厚朴中总酚的含量较前期增长了310％,环割处理的凹叶厚朴中总酚的含量较前期增加了125．65％,铁丝环扎处理的凹叶厚朴中总酚的含量较前期增加了343．2％,而同期的空白组中凹叶厚朴中总酚的含量较前期增加了84．62％。实验结果显示不同栽培措施的效果由大到小的顺序是：铁丝环扎〉修枝〉环割。结论：凹叶厚朴中厚朴酚、和厚朴酚含量的积累可以通过改变土壤的肥力、环割、修枝等栽培措施得以提高。     

Invariance priors for Bayesian feed-forward neural networks

Neural networks (NN) are famous for their advantageous flexibility for problems when there is insufficient knowledge to set up a proper model. On the other hand, this flexibility can cause overfitting and can hamper the generalization of neural networks. Many approaches to regularizing NN have been suggested but most of them are based on ad hoc arguments. Employing the principle of transformation invariance, we derive a general prior in accordance with the Bayesian probability theory for feed-forward networks. An optimal network is determined by Bayesian model comparison, verifying the applicability of this approach. Additionally the prior presented affords cell pruning.     

Changes in the second messenger cyclic AMP during development may underlie motoric symptoms in attention deficit/hyperactivity disorder (ADHD)

The transitions that occur in the ascending dopamine systems between childhood and adulthood parallel the emergence, course, and severity of attention-deficit hyperactivity disorder (ADHD) symptoms. Behaviorally, rats are more active in open field during periadolescence, and activity levels decline by 50% in males by adulthood. This peak in behavior parallels a transient overproduction in D1 and D2 dopamine receptors that occurs at puberty in rat striatum (STR) and prefrontal cortex (PFC), followed by a decline in receptor density into adulthood. While tempting to speculate that receptor density plays a role in the waning of ADHD symptoms, receptor overproduction does not occur in the nucleus accumbens (NA), which demonstrates only a modest rise in receptor density (10–20%). Given the importance of the accumbens in locomotor activity, an alternative explanation for increased activity was sought. The second messenger system cyclic adenosine monophosphate (cAMP) has classically been associated with dopamine receptors. The results of these studies demonstrate that cAMP accumulation in the accumbens and the STR parallel the observed rise and fall in activity levels in rats. At puberty, basal cAMP levels are 35% higher relative to adulthood in male accumbens, while a modest 7% change was observed in STR. Forskolin-stimulated cAMP was 240–300% higher in STR and accumbens at puberty before declining with maturation. These findings suggest that, the adolescent dopamine system has a much higher ‘tone’ relative to adults. However, pharmacological responsiveness of cAMP to D1 or D2 stimulation demonstrates an overall blunted response during puberty relative to adulthood. This finding is consistent with a hyposensitivity to stress and pharmacological agents at puberty in animals that are in a hyperdopaminergic state. These findings of combined elevated cAMP accumulation and reduced cAMP sensitivity during adolescence have clinical implications for hypothesized mechanism and course of ADHD and its treatment. The maturational decline in cAMP activity may explain why this disorder recedes, while, simultaneously cAMP becomes more responsive to D1 and D2 receptor stimulation in adulthood.     

Consequences of pruning in morphometry of coronary vasculature

There is a paucity of data on the branching pattern and dimensions of the blood vessels in various organs. The reason for the paucity is undoubtedly the tremendous effort needed to obtain the morphometric data. For those organs whose morphometric data have been determined, pruning was introduced: cutting off branches at succesive generations, measuring what remained, using the statistical data to estimate what were cut off, and adding the estimated data to the measured data to obtain the final results. Evaluation of the effects of pruning was not possible, however, because a full set of precise data did not exist. Now a complete set of morphometric data on the coronary arteries is presented by Kassabet al. (8). Hence we are in a position to evaluate pruningversus accuracy. Among several pruning protocols tried we found a simple, easy-to-follow scheme that seemed to be reasonable. It reduced the labor by 79% when it was applied to the left anterior descending (LAD) artery of the pig, and it caused the following percentage errors based on comparison with the unpruned data. The largest error incurred in the mean diameters of all orders of tree is 7.6%. The corresponding maximum errors in the length and number of elements in all orders are −9.8% and 30.0%, respectively. The estimated error of the total equivalent Poiseuille's resistance for the LAD artery computed from pruned data was 25.2% when compared with that computed from unpruned data.     

Factoring neurotrophins into a neurite-based pathophysiological model of schizophrenia

Neurotrophins are growth factors that, through variations in concentration and changes in receptor expression, regulate the formation of axons and dendrites during development and throughout adult life. Here we review these growth factors, particularly in the context of schizophrenia, a psychiatric disorder characterized by neurodevelopmental abnormalities. We first discuss emerging information derived from physiologically relevant organotypic cultures and in vivo studies regarding the effects of neurotrophins on the neuronal structure including pruning and GABAergic neurons. We then review postmortem studies of neurotrophin levels and their receptors in brains of individuals with schizophrenia, and compare them with what is known about neurotrophin effects on neuronal structure. This comparison indicates that only some neuropathological defects encountered in patients with schizophrenia can be explained by the single action of neurotrophins on dendrites and axons. However, we propose that a number of inconsistent findings and apparently unrelated results in the schizophrenia field can be reconciled if neurons are considered structurally plastic cells capable of extending and retracting dendrites and axons throughout life.     

The POP learning algorithms: reducing work in identifying fuzzy rules

A novel fuzzy neural network, the Pseudo Outer-Product based Fuzzy Neural Network (POPFNN), and its two fuzzy-rule-identification algorithms are proposed in this paper. They are the Pseudo Outer-Product (POP) learning and the Lazy Pseudo Outer-Product (LazyPOP) learning algorithms. These two learning algorithms are used in POPFNN to identify relevant fuzzy rules. In contrast with other rule-learning algorithms, the proposed algorithms have many advantages, such as being fast, reliable, efficient, and easy to understand. POP learning is a simple one-pass learning algorithm. It essentially performs rule-selection. Hence, it suffers from the shortcoming of having to consider all the possible rules. The second algorithm, the LazyPOP learning algorithm, truly identifies the fuzzy rules which are relevant and does not use a rule-selection method whereby irrelevant fuzzy rules are eliminated from an initial rule set. In addition, it is able to adjust the structure of the fuzzy neural network. The proposed LazyPOP learning algorithm is able to delete invalid feature inputs according to the fuzzy rules that have been identified. Extensive experimental results and discussions are presented for a detailed analysis of the proposed algorithms.