Artificial neural networks(ANN) have been extensively researched due to their significant energy-saving benefits.Hardware implementations of ANN with dropout function would be able to avoid the overfitting problem. Th...Artificial neural networks(ANN) have been extensively researched due to their significant energy-saving benefits.Hardware implementations of ANN with dropout function would be able to avoid the overfitting problem. This letter reports a dropout neuronal unit(1R1T-DNU) based on one memristor–one electrolyte-gated transistor with an ultralow energy consumption of 25 p J/spike. A dropout neural network is constructed based on such a device and has been verified by MNIST dataset, demonstrating high recognition accuracies(> 90%) within a large range of dropout probabilities up to40%. The running time can be reduced by increasing dropout probability without a significant loss in accuracy. Our results indicate the great potential of introducing such 1R1T-DNUs in full-hardware neural networks to enhance energy efficiency and to solve the overfitting problem.展开更多
Previous studies have shown that models of depression exhibit structural and functional changes to the neurovascular unit. Thus, we hypothesized that diabetes-related depression might be associated with damage to the ...Previous studies have shown that models of depression exhibit structural and functional changes to the neurovascular unit. Thus, we hypothesized that diabetes-related depression might be associated with damage to the hippocampal neurovascular unit. To test this hypothesis, neurons, astrocytes and endothelial cells were isolated from the brain tissues of rat embryos and newborn rats. Hippocampal neurovascular unit co-cultures were produced using the Transwell chamber co-culture system. A model of diabetes-related depression was generated by adding 150 mM glucose and 200 μM corticosterone to the culture system and compared with the neuron + astrocyte and astrocyte + endothelial cell co-culture systems. Western blot assay was used to measure levels of structural proteins in the hippocampal neurovascular unit co-culture system. Levels of basic fibroblast growth factor, angiogenic factor 1, glial cell line–derived neurotrophic factor, transforming growth factor β1, leukemia inhibitory factor and 5-hydroxytryptamine in the hippocampal neurovascular unit co-culture system were measured by enzyme-linked immunosorbent assay. Flow cytometry and terminal deoxynucleotidyl transferase(TdT)-mediated dUTP nick end labeling staining was used to assess neuronal apoptosis in the hippocampal neurovascular unit. The neurovascular unit triple cell co-culture system had better barrier function and higher levels of structural and secretory proteins than the double cell co-culture systems. In comparison, in the model of diabetes-related depression, the neurovascular unit was damaged with decreased barrier function, poor structural integrity and impaired secretory function. Moreover, neuronal apoptosis was markedly increased, and 5-hydroxytryptamine levels were reduced. These results suggest that diabetes-related depression is associated with structural and functional damage to the neurovascular unit. Our findings provide a foundation for further studies on the pathogenesis of diabetes-related depression.展开更多
Non-adherent bone marrow cell-derived mesenchymal stem cells from C57BL/6J mice were sepa- rated and cultured using the ＂pour-off＂ method. Non-adherent bone marrow cell-derived mesen- chymal stem ceils developed col...Non-adherent bone marrow cell-derived mesenchymal stem cells from C57BL/6J mice were sepa- rated and cultured using the ＂pour-off＂ method. Non-adherent bone marrow cell-derived mesen- chymal stem ceils developed colony-forming unit-fibroblasts, and could be expanded by supple- mentation with epidermal growth factor. Immunocytochemistry showed that the non-adherent bone marrow cell-derived mesenchymal stem cells exposed to basic fibroblast growth factor/epidermal growth factor/nerve growth factor expressed the neuron specific markers, neurofilament-200 and NeuN, in vitro. Non-adherent bone marrow cell-derived mesenchymal stem cells from 13-galactosidase transgenic mice were also transplanted into focal ischemic brain （right corpus striatum） of C57BL/6J mice. At 8 weeks, cells positive for LacZ and 13-galactosidase staining were observed in the ischemic tissues, and cells co-labeled with both 13-galactosidase and NeuN were seen by double immunohistochemical staining. These findings suggest that the non-adherent bone marrow cell-derived mesenchymal stem cells could differentiate into neuronal-like cells in vitro and in vivo.展开更多
BACKGROUND：Amyotrophic lateral sclerosis （ALS） is the most common of all the motor neuron diseases and the absence of a biologic marker has made both diagnosis and tracking evolution of the disease difficult, Elect...BACKGROUND：Amyotrophic lateral sclerosis （ALS） is the most common of all the motor neuron diseases and the absence of a biologic marker has made both diagnosis and tracking evolution of the disease difficult, Electrodiagnostic tests play a fundamental role in quantifying pathological changes in the motor unit pool.OBJECTIVE：We assessed distal-proximal Motor Unit （MU） loss and changes using the method of motor unit number estimation （MUNE）.DESIGN, TIME AND SETTING：A case-control study was performed at the Department of Neuroscience, Pisa University Medical School, Italy from December 1999 to November 2009. PARTICIPANTS：A total of 50 ALS patients were recruited, 30 males：mean age （59.6 ± 13.3） years; 20 females：mean age （63.9 ± 11.7） years; range （30-82） years; all patients had probable or definite ALS. Thirty healthy volunteers were recruited from department staffs, including 20 males and 10 females; mean age （57.7 ± 13.8） years served as controls.METHODS：MUNE was performed for both the biceps brachii and abductor digiti minimi muscles of the same side. The technique used relayed substantially on manual incremental stimulation of the motor nerve, known as the McComas technique （50 ms sweep duration, a gain of 2 mV/Div for M wave, 0.5 mV/Div for each step; filters 10-20 kHz）.MAIN OUTCOME MEASURES：MUNE results were measured.RESULTS：Functioning MU numbers, measured by MUNE, decreased in the biceps brachii and abductor digiti minimi muscles over the entire one-year follow-up period （one assessment every three months） compared with baseline determination, the rate of MU decrease was similar in both muscles, but steeper distally.CONCLUSION：MUNE is a feasible method for ALS patients both proximally and distally to track changes over time in muscle MUs during the disease＇s evolution.展开更多
The scope of this paper is to forecast wind speed. Wind speed, temperature, wind direction, relative humidity, precipitation of water content and air pressure are the main factors make the wind speed forecasting as a ...The scope of this paper is to forecast wind speed. Wind speed, temperature, wind direction, relative humidity, precipitation of water content and air pressure are the main factors make the wind speed forecasting as a complex problem and neural network performance is mainly influenced by proper hidden layer neuron units. This paper proposes new criteria for appropriate hidden layer neuron unit’s determination and attempts a novel hybrid method in order to achieve enhanced wind speed forecasting. This paper proposes the following two main innovative contributions 1) both either over fitting or under fitting issues are avoided by means of the proposed new criteria based hidden layer neuron unit’s estimation. 2) ELMAN neural network is optimized through Modified Grey Wolf Optimizer (MGWO). The proposed hybrid method (ELMAN-MGWO) performance, effectiveness is confirmed by means of the comparison between Grey Wolf Optimizer (GWO), Adaptive Gbest-guided Gravitational Search Algorithm (GGSA), Artificial Bee Colony (ABC), Ant Colony Optimization (ACO), Cuckoo Search (CS), Particle Swarm Optimization (PSO), Evolution Strategy (ES), Genetic Algorithm (GA) algorithms, meanwhile proposed new criteria effectiveness and precise are verified comparison with other existing selection criteria. Three real-time wind data sets are utilized in order to analysis the performance of the proposed approach. Simulation results demonstrate that the proposed hybrid method (ELMAN-MGWO) achieve the mean square error AVG ± STD of 4.1379e-11 ± 1.0567e-15, 6.3073e-11 ± 3.5708e-15 and 7.5840e-11 ± 1.1613e-14 respectively for evaluation on three real-time data sets. Hence, the proposed hybrid method is superior, precise, enhance wind speed forecasting than that of other existing methods and robust.展开更多
基金Project supported by the National Key Research and Development Program of China (Grant Nos. 2021YFA1202600 and 2023YFE0208600)in part by the National Natural Science Foundation of China (Grant Nos. 62174082, 92364106, 61921005, 92364204, and 62074075)。
文摘Artificial neural networks(ANN) have been extensively researched due to their significant energy-saving benefits.Hardware implementations of ANN with dropout function would be able to avoid the overfitting problem. This letter reports a dropout neuronal unit(1R1T-DNU) based on one memristor–one electrolyte-gated transistor with an ultralow energy consumption of 25 p J/spike. A dropout neural network is constructed based on such a device and has been verified by MNIST dataset, demonstrating high recognition accuracies(> 90%) within a large range of dropout probabilities up to40%. The running time can be reduced by increasing dropout probability without a significant loss in accuracy. Our results indicate the great potential of introducing such 1R1T-DNUs in full-hardware neural networks to enhance energy efficiency and to solve the overfitting problem.
基金supported by the National Natural Science Foundation of China,No.81373578(to YHW),81573965(to YHW)the Natural Science Foundation of Hunan Province of China,No.2017JJ3241(to JL)the Education Department Scientific Research Foundation of Hunan Province of China,No.17C1229(to JL)
文摘Previous studies have shown that models of depression exhibit structural and functional changes to the neurovascular unit. Thus, we hypothesized that diabetes-related depression might be associated with damage to the hippocampal neurovascular unit. To test this hypothesis, neurons, astrocytes and endothelial cells were isolated from the brain tissues of rat embryos and newborn rats. Hippocampal neurovascular unit co-cultures were produced using the Transwell chamber co-culture system. A model of diabetes-related depression was generated by adding 150 mM glucose and 200 μM corticosterone to the culture system and compared with the neuron + astrocyte and astrocyte + endothelial cell co-culture systems. Western blot assay was used to measure levels of structural proteins in the hippocampal neurovascular unit co-culture system. Levels of basic fibroblast growth factor, angiogenic factor 1, glial cell line–derived neurotrophic factor, transforming growth factor β1, leukemia inhibitory factor and 5-hydroxytryptamine in the hippocampal neurovascular unit co-culture system were measured by enzyme-linked immunosorbent assay. Flow cytometry and terminal deoxynucleotidyl transferase(TdT)-mediated dUTP nick end labeling staining was used to assess neuronal apoptosis in the hippocampal neurovascular unit. The neurovascular unit triple cell co-culture system had better barrier function and higher levels of structural and secretory proteins than the double cell co-culture systems. In comparison, in the model of diabetes-related depression, the neurovascular unit was damaged with decreased barrier function, poor structural integrity and impaired secretory function. Moreover, neuronal apoptosis was markedly increased, and 5-hydroxytryptamine levels were reduced. These results suggest that diabetes-related depression is associated with structural and functional damage to the neurovascular unit. Our findings provide a foundation for further studies on the pathogenesis of diabetes-related depression.
基金supported by the National Natural Science Foundation of China,No.30471836
文摘Non-adherent bone marrow cell-derived mesenchymal stem cells from C57BL/6J mice were sepa- rated and cultured using the ＂pour-off＂ method. Non-adherent bone marrow cell-derived mesen- chymal stem ceils developed colony-forming unit-fibroblasts, and could be expanded by supple- mentation with epidermal growth factor. Immunocytochemistry showed that the non-adherent bone marrow cell-derived mesenchymal stem cells exposed to basic fibroblast growth factor/epidermal growth factor/nerve growth factor expressed the neuron specific markers, neurofilament-200 and NeuN, in vitro. Non-adherent bone marrow cell-derived mesenchymal stem cells from 13-galactosidase transgenic mice were also transplanted into focal ischemic brain （right corpus striatum） of C57BL/6J mice. At 8 weeks, cells positive for LacZ and 13-galactosidase staining were observed in the ischemic tissues, and cells co-labeled with both 13-galactosidase and NeuN were seen by double immunohistochemical staining. These findings suggest that the non-adherent bone marrow cell-derived mesenchymal stem cells could differentiate into neuronal-like cells in vitro and in vivo.
基金Supported by the Italian MIUR PRIN Grant year 2006,# 2006062332_002
文摘BACKGROUND：Amyotrophic lateral sclerosis （ALS） is the most common of all the motor neuron diseases and the absence of a biologic marker has made both diagnosis and tracking evolution of the disease difficult, Electrodiagnostic tests play a fundamental role in quantifying pathological changes in the motor unit pool.OBJECTIVE：We assessed distal-proximal Motor Unit （MU） loss and changes using the method of motor unit number estimation （MUNE）.DESIGN, TIME AND SETTING：A case-control study was performed at the Department of Neuroscience, Pisa University Medical School, Italy from December 1999 to November 2009. PARTICIPANTS：A total of 50 ALS patients were recruited, 30 males：mean age （59.6 ± 13.3） years; 20 females：mean age （63.9 ± 11.7） years; range （30-82） years; all patients had probable or definite ALS. Thirty healthy volunteers were recruited from department staffs, including 20 males and 10 females; mean age （57.7 ± 13.8） years served as controls.METHODS：MUNE was performed for both the biceps brachii and abductor digiti minimi muscles of the same side. The technique used relayed substantially on manual incremental stimulation of the motor nerve, known as the McComas technique （50 ms sweep duration, a gain of 2 mV/Div for M wave, 0.5 mV/Div for each step; filters 10-20 kHz）.MAIN OUTCOME MEASURES：MUNE results were measured.RESULTS：Functioning MU numbers, measured by MUNE, decreased in the biceps brachii and abductor digiti minimi muscles over the entire one-year follow-up period （one assessment every three months） compared with baseline determination, the rate of MU decrease was similar in both muscles, but steeper distally.CONCLUSION：MUNE is a feasible method for ALS patients both proximally and distally to track changes over time in muscle MUs during the disease＇s evolution.
文摘The scope of this paper is to forecast wind speed. Wind speed, temperature, wind direction, relative humidity, precipitation of water content and air pressure are the main factors make the wind speed forecasting as a complex problem and neural network performance is mainly influenced by proper hidden layer neuron units. This paper proposes new criteria for appropriate hidden layer neuron unit’s determination and attempts a novel hybrid method in order to achieve enhanced wind speed forecasting. This paper proposes the following two main innovative contributions 1) both either over fitting or under fitting issues are avoided by means of the proposed new criteria based hidden layer neuron unit’s estimation. 2) ELMAN neural network is optimized through Modified Grey Wolf Optimizer (MGWO). The proposed hybrid method (ELMAN-MGWO) performance, effectiveness is confirmed by means of the comparison between Grey Wolf Optimizer (GWO), Adaptive Gbest-guided Gravitational Search Algorithm (GGSA), Artificial Bee Colony (ABC), Ant Colony Optimization (ACO), Cuckoo Search (CS), Particle Swarm Optimization (PSO), Evolution Strategy (ES), Genetic Algorithm (GA) algorithms, meanwhile proposed new criteria effectiveness and precise are verified comparison with other existing selection criteria. Three real-time wind data sets are utilized in order to analysis the performance of the proposed approach. Simulation results demonstrate that the proposed hybrid method (ELMAN-MGWO) achieve the mean square error AVG ± STD of 4.1379e-11 ± 1.0567e-15, 6.3073e-11 ± 3.5708e-15 and 7.5840e-11 ± 1.1613e-14 respectively for evaluation on three real-time data sets. Hence, the proposed hybrid method is superior, precise, enhance wind speed forecasting than that of other existing methods and robust.
