Development of a recurrent Sigma-Pi neural network rainfall forecasting system in Hong Kong

At the moment, weather forecasting is still an art — the experience and intuition of forecasters play a significant role in determining the quality of forecasting. This paper describes the development of a new approach to rainfall forecasting using neural networks. It deals with the extraction of information from radar images and an evaluation of past rain gauge records to provide shortterm rainfall forecasting. All of the meteorological data were provided by the Royal Observatory of Hong Kong (ROHK). Preprocessing procedures were essential for this neural network rainfall forecasting. The forecast of the rainfall was performed every half an hour so that a storm warning signal can be delivered to the public in advance. The network architecture is based on a recurrent Sigma-Pi network. The results are very promising, and this neural-based rainfall forecasting system is capable of providing a rain storm warning signal to the Hong Kong public one hour ahead.     

Prevalence of adrenal carcinoma among incidentally discovered adrenal masses. A retrospective study from 1989 to 1994. Gruppo Piemontese Incidentalomi Surrenalici

BACKGROUND: The incidental discovery of an adrenal mass poses the problem of distinguishing between the frequent benign masses and the infrequent malignant ones that require surgery. Univocal guidelines to approach this problem are unavailable. OBJECTIVE: To perform a survey of the clinical management of incidentally discovered adrenal masses (ie, adrenal incidentalomas). DESIGN: A multicentric retrospective analysis of hospital medical records of adrenal incidentalomas diagnosed during a 5-year period; the medical records were scrutinized for demographic data and clinical details by means of a specifically tailored questionnaire. SETTING: The major surgical and medical centers of Piedmont, a northern Italian region with approximately 4 million inhabitants. The recruitment pattern of these centers was unselected. PATIENTS: The definition of adrenal incidentaloma was limited to patients with a physical examination and a clinical history unindicative of adrenal disease. Exclusion criteria also included hypertension of suspected endocrine origin and a history of neoplasms known to metastasize frequently in the adrenal glands. Two hundred twenty-four medical records were collected, and 210 were analyzed (14 excluded a posteriori). RESULTS: Most patients were in their 50s and 60s, and women were predominantly affected. The frequency of adrenocortical cancer was 13% among patients operated on. The tumor diameter was highly correlated with the risk of cancer; a cutoff at 5 cm had a sensitivity of 93% with a specificity of 64% in discriminating between benign and malignant cortical lesions. CONCLUSIONS: The occurrence of adrenocortical carcinoma among adrenal incidentalomas is not rare. The evaluation of the mass size is a simple and effective method for selecting patients at risk for cancer. The indication for surgery of masses larger than 5 cm, or of masses of any diameter that have suspicious imaging characteristics, limits unnecessary operations and costs.     

A Metallographic Study of Abnormal Grain Growth During Aging of a β-Quenched Zr-1.14 Wt% Cr-0.08 Wt% Fe Alloy

Results are reported of a detailed optical and electron metallographic study of the effects of variations in quenching rate from the β-phase and aging temperature in the + ZrCr₂ phase region, on the recrystallization and grain growth characteristics of Zr-1.14 wt% Cr-0.08 wt% Fe alloy used for nuclear fuel sheathing. It is shown that certain quenching rates, which produce a mixed martensitic-Widmanstätten structure, should be avoided during the fabrication of fuel sheathing, since recrystallization and grain growth occurs at a much lower temperature than a material quenched at such a rate that the structure is wholly Widmanstätten or wholly martensitic. Preaging at 350 or 450°C prior to the high-temperature (700–750°C) aging reduces the incidence of recrystallization and grain growth.     

Direct evidence of high DNA binding activity of transcription factor AP-1 in rheumatoid arthritis synovium

Green fluorescent protein (GFP) is increasingly being used in plant biology from the cellular level to whole plant level. At the cellular level, GFP is being used as an in vivo reporter to assess frequency of transient and stable transformation. GFP has also proven to be an invaluable tool in monitoring trafficking and subcellular localization of protein. At the organ level and up, many exciting applications are rapidly emerging. The development of brighter GFP mutants with more robust folding properties has enabled better macroscopic visualization of GFP in whole leaves and plants. One interesting example has been the use of GFP to monitor virus movement in and among whole plants. GFP is also emerging as a powerful tool to monitor transgene movement and transgenic plants in the field. In a proof-of-concept study, tobacco was transformed with a modified version of the GFP gene controlled by a constitutive (35S) promoter. GFP expression in progeny plants ranged from 0% to 0.5%, and approximately 0.1% GFP was the minimal amount needed for unambiguous macroscopic detection. GFP is the first truly in vivo reporter system useful in whole plants, and we project its usefulness will increase even further as better forms of GFP genes become available.     

Artificial neural networks for coordination and control: The portability of experiential representations

It is time to locate connectionist representation theory in the new wave of robotics research. The utility of representations developed in artificial neural networks (ANNs) during learning has been demonstrated in cognitive science research since the 1980s. The research reported here puts learned representations to work in a decentered control task, the disembodied arm problem, in which a mobile robot operates an arm fixed to a table to pick up objects. There is no physical linkage between the arm and the robot and so the robot's point of view must be decentered. This is done by developing a modular Artificial Neural Net system in three stages: (i) a classifier net is trained with laser scan data to output transformationally invariant position classes; (ii) an arm net is trained for picking up objects; (iii) an inter net is trained to communicate and coordinate the sensing and acting. The completed system is shown to create new nonsymbolic transformationally invariant representations in order to perform the effective generalization of decentered viewpoints.     

Magnetic resonance of the encephalon in 17 patients with ocular Beh?et''s disease

OBJECTIVE: To evaluate the relationships between patient and physician pretreatment expectations of pain relief and subsequent pain relief reported by chronic pain patients immediately after treatment. DESIGN: Prospective study of consecutive patients undergoing a procedure in a pain clinic for treatment of chronic pain. Patients rated their current pain level and their expectation of pain relief immediately prior to undergoing a procedure (e.g., intravenous drug infusion, nerve block) for the treatment of chronic pain. Simultaneously and independently, the treating physician completed a similar questionnaire. At completion of the procedure, patients rated their current pain level and degree of pain relief. SETTING: University of Washington Multidisciplinary Pain Center procedure suite. PATIENTS: Forty-six consecutive chronic pain patients. INTERVENTION: Intravenous drug infusions and nerve blocks. OUTCOME MEASURES: Current pain and pain relief ratings. RESULTS: Patients' pain relief expectation ratings were not correlated significantly with their postprocedure pain relief ratings or pre-post procedure changes in pain ratings. However, a statistically significant correlation was found between physician expectations of pain relief and patient pain relief ratings and patient pre-post procedure changes in pain. CONCLUSIONS: The results of this study suggest that physicians are better predictors than are patients of patients responses to these procedures and/or that physicians may somehow subtly communicate their expectations to patients during the procedure, and these expectations then influence patient response. Patient pretreatment expectations may not always play a significant role in nonspecific treatment effects.     

Neurochemical studies of schizophrenia in Japan

Neurobiological research in schizophrenia has been hampered by several confounding factors such as the heterogeneity of the illness and the paucity of biological markers. Recent progress in research methods, however, has enabled the improvement in our understanding its pathophysiology. This paper reviews recent neurochemical investigations of schizophrenia and its animal models which were conducted in Japan in the last decade. The research areas reviewed are (i) monoamine and their metabolites in body fluids, (ii) phospholipids and prostaglandins, (iii) neurochemistry in autopsy brains, (iv) immunological measures, (v) magnetic resonance spectroscopy, (vi) regional cerebral blood flows (rCBF), (vii) molecular genetics, and (viii) animal models. It is worth noting that there exist abnormalities of amino acidergic (glutamatergic and GABAergic) neurotransmission as well as monoaminergic (dopaminergic and serotonergic) one in postmortem schizophrenic brains. These abnormalities and also the findings of altered rCBF indicate the existence of disturbed neuronal circuits that contribute to the diverse symptoms of schizophrenia. Also, dysfunction of membrane phospholipids derived from studies on magnetic resonance spectroscopy may underlie negative symptoms in schizophrenia. Given that schizophrenia is considered to comprise a group of disorders with a diverse heterogeneity of etiologies, research in the next decade is expected to identify putative genes that are involved in vulnerability to schizophrenic phenotype.     

Analysis of the antidiabetic drug acarbose by capillary electrophoresis

This study describes the derivatization of the pseudooligosaccharide acarbose and its main metabolite, component 2, with 7-aminonaphthalene-1,3-disulfonic acid (ANDS) in human urine. Their efficient separation was possible by means of capillary zone electrophoresis, using a capillary tube of fused-silica containing 100 mM triethylammonium phosphate buffer, pH 1.5. On column laser-induced fluorescence allowed the detection of the pseudooligosaccharides in human urine in the nanomolar range. With this method, acarbose and component 2 were quantified in human urine after application of 300 mg of acarbose.     

Molecular medicine and population control

As we uncover the physiological and molecular mechanisms behind human reproduction, we gain the potential to exert more control over our reproductive capabilities. In the past two decades, the prospects for 'infertile' women to bear children, or 'sterile' men to father them, have improved dramatically; recently, women have given birth in their sixties, well beyond their natural menopause. Thanks to developments in contraception, couples can now enjoy an active sex-life without a significant risk of pregnancy, and more accurately control the size and timing of their families. But these new freedoms have a price: they also have the potential to cause demographic distortions, medical harm to individuals and abuse of human rights.     

Certain nonlinear partially observable stochastic optimal controlproblems with explicit control laws equivalent to LEQG/LQG problems

This paper is concerned with partially observed stochastic optimal control problems when nonlinearities enter the dynamics of the unobservable state and the observations as gradients of potential functions. Explicit representations for the information state are derived in terms of a finite number of sufficient statistics. Consequently, the partially observed problem is recast as one of complete information with a new state generated by a modified version of the Kalman filter. When the terminal cost is quadratic in the unobservable state and includes the integral of the nonlinearities, the optimal control laws are explicitly computed, similar to linear-exponential-quadratic-Gaussian (LEQG) and linear-quadratic-Gaussian (LQG) tracking problems. The results are applicable to filtering and control of Hamiltonian systems