Multiple Network-on-Chip Model for High Performance Neural Network

Dong, Yiping,Li, Ce,Lin, Zhen,Watanabe, Takahiro The Institute of Electronics and Information Engin 2010 Journal of semiconductor technology and science Vol.10 No.1

Hardware implementation methods for Artificial Neural Network (ANN) have been researched for a long time to achieve high performance. We have proposed a Network on Chip (NoC) for ANN, and this architecture can reduce communication load and increase performance when an implemented ANN is small. In this paper, a multiple NoC models are proposed for ANN, which can implement both a small size ANN and a large size one. The simulation result shows that the proposed multiple NoC models can reduce communication load, increase system performance of connection-per-second (CPS), and reduce system running time compared with the existing hardware ANN. Furthermore, this architecture is reconfigurable and reparable. It can be used to implement different applications of ANN.

Low Power and High Speed Network on Chip Architecture for BP Neural Network

Yiping Dong,Yinghe Li,Yang Wang,Takahiro Watanabe 대한전자공학회 2009 ITC-CSCC :International Technical Conference on Ci Vol.2009 No.7

Back Propagation Artificial Neural Network (BP-ANN), one of widely used neural networks, has been used in a lot of areas. Hardware architecture of BP-ANN was proposed by using an NoC (Network on Chip) and implemented on FPGA. In this paper, new NoC architecture which has a torus topology and IXY (Intelligent XY) routing algorithm is developed for BP-ANN to make it low power and high performance. This system is implemented by FPGA to estimate system performance and power consumption. NIRGAM NoC simulator is also used to evaluate latency and throughput of this system. Experimental results show that our proposed architecture can increase Connection Per Second (CPS) about 2 times than existing digital hardware ANN; it can reduce communication load which total packet size can reduce about 3.2 times compared with traditional packet transmit method of BP-ANN. It can reduce latency by 23.7% and dynamic power consumption by 7.2% compare with the former NoC architecture BP-ANN. It is reconfigurable and expandable to meet various ANN applications. Furthermore other type ANNs can also be implemented in the system by adjusting a routing algorithm of NoCs.

Network on Chips Structure for Mapping Two Hidden Layers BP-ANNs

Yiping, Dong,Takahiro, Watanabe 대한전자공학회 2008 ITC-CSCC :International Technical Conference on Ci Vol.2008 No.7

Artificial neural networks (ANNs) are widely used in various applications such as recognition, security, computer learning and so on. A two hidden layers Back-Propagation ANN (BP-ANN) is a typical one. But BP-ANNs are hard to be implemented in hardware due to interconnections and cost. Recently, Networkson-Chips (NoCs) where a packet-based network is used routing on-chip signals, have begun to attract as a smart structure to solve the interconnect problem. It can also map one or more logical units into a single physical hardware unit and connect with on router. In this paper we propose a system with NoCs structure to mapping 5 neurons in one router for the two hidden layers BP-ANNs. Our system is evaluated for the latency and throughput using NIRGAM NoCs simulator, and is implemented on an FPGA device to estimate system performance and power consumption. Experimental results show that our proposed system has a great reduction in communication load, low latency and a high throughput. It is also reconfigurable and expandable to meet various NN application problems, and besides, not only BP-ANNs but also a random-connected ANNs or any type ANNs can be implemented in the system by adjusting a routing algorithm of NoC.

Multiple Network-on-Chip Model for High Performance Neural Network

Yiping Dong,Ce Li, Zhen Lin,Takahiro Watanabe 대한전자공학회 2010 Journal of semiconductor technology and science Vol.10 No.1

Hardware implementation methods for Artificial Neural Network (ANN) have been researched for a long time to achieve high performance. We have proposed a Network on Chip (NoC) for ANN, and this architecture can reduce communication load and increase performance when an implemented ANN is small. In this paper, a multiple NoC models are proposed for ANN, which can implement both a small size ANN and a large size one. The simulation result shows that the proposed multiple NoC models can reduce communication load, increase system performance of connection-per-second (CPS), and reduce system running time compared with the existing hardware ANN. Furthermore, this architecture is reconfigurable and reparable. It can be used to implement different applications of ANN.

Self-Rated Health Status and Risk of Incident Stroke in 0.5 Million Chinese Adults: The China Kadoorie Biobank Study

Wenhong Dong,Xiong-Fei Pan,Canqing Yu,Jun Lv,Yu Guo,Zheng Bian,Ling Yang,Yiping Chen,Tangchun Wu,Zhengming Chen,An Pan,Liming Li 대한뇌졸중학회 2018 Journal of stroke Vol.20 No.2

Background and purpose Self-rated health (SRH) is a consistent and strong predictor of all-cause and cardiovascular mortality in various populations. However, the associations between SRH measures and risk of first-ever or recurrent stroke were rarely explored. We thus aim to prospectively investigate the associations between SRH measures and risk of total and subtypes of stroke in Chinese population. Methods A total of 494,113 participants from the China Kadoorie Biobank without prior heart diseases or cancer (486,541 without stroke and 7,572 with stroke) were followed from baseline (2004 to 2008) until December 31, 2013. General and age-comparative SRH were obtained from baseline questionnaires. First-ever stroke or recurrent events were ascertained through linkage to disease registry system and health insurance data. Results We identified 27,662 first-ever stroke and 2,909 recurrent events during an average of 7.0 years of follow-up. Compared with excellent general SRH, the hazard ratios (HRs) and 95% confidence intervals (CIs) for first-ever stroke associated with good, fair, and poor general SRH were 1.04 (1.00 to 1.08), 1.19 (1.15 to 1.23), and 1.49 (1.42 to 1.56) in the multivariate model, respectively. Compared with better age-comparative SRH, the HRs (95% CIs) of same and worse age-comparative SRH were 1.13 (1.10 to 1.17) and 1.51 (1.45 to 1.58), respectively. The relations of SRH measures with ischemic stroke, hemorrhagic stroke, and recurrent stroke were similar to that with total first-ever stroke. However, the magnitude of associations was much stronger for fatal stroke than for non-fatal stroke. Conclusions This large-scale prospective cohort suggests that self-perceived health status is associated with incident stroke, regardless of stroke subtype.

One-step multiplexed detection of foodborne pathogens: Combining a quantum dot-mediated reverse assaying strategy and magnetic separation

Yin, Binfeng,Wang, Yu,Dong, Mingling,Wu, Jing,Ran, Bei,Xie, Mengxia,Joo, Sang Woo,Chen, Yiping Elsevier 2016 Biosensors & bioelectronics Vol.86 No.-

<P><B>Abstract</B></P> <P>A rapid and multiplexed immunosensor was developed based on a quantum dot (QD)-reverse assaying strategy (RAS) and immuno-magnetic beads (IMBs) for one-step and simultaneous detection of <I>Escherichia coli O157: H7</I> and <I>Salmonella</I>. In a conventional QD-based immunosensor, the fluorescence signal of the “IMBs-target-QD” immunoconjugate is directly used as the assaying readout. However, the fluorescence signal is affected by IMBs due to light scattering and the “IMBs-target-QD” immunoconjugate needs multiple washing and re-suspension steps. To address these problems, we use the surplus QD-antibody conjugate as signal readout in the RAS, which prevents interference from the IMBs, increases the fluorescence signal, and avoids complex operations. Compared with conventional QD-based immunosensor, the sensitivity of QD-RSA immunosensor for detection of <I>Escherichia coli O157: H7</I> has been improved fifty-fold, and whole analysis procedure can be finished within 1h. Therefore, this RSA strategy is promising for improving the performance of QD-based immunosensors and could greatly broaden their applications.</P> <P><B>Highlights</B></P> <P> <UL> <LI> The immunosensor based on quantum dot-reverse assaying strategy (QD-RAS) was developed. </LI> <LI> The sensitivity of QD-RAS improved 50 times compared to conventional QDs-immunosensor. </LI> <LI> The whole analysis in QD-RAS can be finished in one-step within 1h. </LI> </UL> </P>

An enzyme-mediated competitive colorimetric sensor based on Au@Ag bimetallic nanoparticles for highly sensitive detection of disease biomarkers

Yin, Binfeng,Zheng, Wenshu,Dong, Mingling,Yu, Wenbo,Chen, Yiping,Joo, Sang Woo,Jiang, Xingyu The Royal Society of Chemistry 2017 The Analyst Vol.142 No.16

<P>We developed a competitive colorimetric nanosensor based on Au@Ag bimetallic nanoparticles for the detection of interleukin-6 (IL-6) in clinical samples. Antibody-conjugated magnetic nanoparticles (MNPs) and polystyrene (PS) microparticles conjugated with both catalase and a secondary antibody can form sandwich structures that enable one-step target enrichment and separation. The catalase on the surface of the PS can catalyze the hydrolysis of hydrogen peroxide (H2O2) to regulate the deposition of Ag<SUP>+</SUP> on the surface of gold nanoparticles (AuNPs), and forms different sizes and amounts of Au@Ag bimetallic nanoparticles (Au@AgNPs) which produce a distinct color signal for readout with the naked eye. Our sensor features high sensitivity, selectivity, reproducibility and anti-interference property as a result of comprehensive parameter optimization. The limit of detection of IL-6 can reach 11 pg mL<SUP>−1</SUP> with the naked eye and 1.2 pg mL<SUP>−1</SUP> by quantitative instrumental analysis. The whole analysis can be finished within 1 h. More importantly, we successfully apply our platform or the detection of IL-6 in clinical samples with better accuracy than conventional enzyme-linked immunosorbent assay (ELISA).</P>

Surface Mesh Generation in Parametric Space Using a Riemannian Surface Definition

Cui Dai,Zhaoxue Wang,Liang Dong,Yiping Chen,Junfeng Qiu 한국유체기계학회 2020 International journal of fluid machinery and syste Vol.13 No.2

In order to solve the problem of generating distortion elements in the mapping from parameter space to real space, and the boundary coincidence of the mesh generated by the software quality, an approach for parametric surface mesh generation based on Riemannian metric, combined with Delaunay triangulation and AFT is proposed. In our algorithm, the boundary curves are discretized based on the proximity and curvature of the curves in the model after derivation the correlation of curve length between parametric space and real space. Background meshes of parametric space were generated by using improved AFT, and could improve the efficient of the algorithm and control element sizing and metric values. When background mesh of parametric space were refined, to counteract mapping distortion, the traditional Delaunay incremental insertion kernel is replaced by inserting the center of triangle circumscribed ellipse, and the algorithm for locating ellipse center and judging whether nodes within ellipse. In this paper, the details of the surface mesh generated by the algorithm are introduced in detail. The algorithm proposed in this paper has the characteristics of reliable algorithm, high mesh generation efficiency and mesh quality. Finally, the reliability of the proposed algorithm is verified by an example of surface mesh generation.

MiRNA320a Inhibitor-Loaded PLGA-PLL-PEG Nanoparticles Contribute to Bone Regeneration in Trauma-Induced Osteonecrosis Model of the Femoral Head

Zhang Ying,Li Chuan,Wei Qiushi,Yuan Qiang,He Wei,Zhang Ning,Dong Yiping,Jing Zhenhao,Zhang Leilei,Wang Haibin,Cao Xiangyang 한국조직공학과 재생의학회 2024 조직공학과 재생의학 Vol.21 No.1

BACKGROUND: This study aimed to explore the effect of a nanomaterial-based miR-320a inhibitor sustained release system in trauma-induced osteonecrosis of the femoral head (TIONFH). METHODS: The miR-320a inhibitor-loaded polyethylene glycol (PEG)- Poly(lactic-co-glycolic acid) (PLGA)- Poly-L-lysine (PLL) nanoparticles were constructed using the double emulsion method. The TIONFH rabbit model was established to observe the effects of miR-320a inhibitor nanoparticles in vivo. Hematoxylin–eosin staining and microcomputed tomography scanning were used for bone morphology analysis. Bone marrow mesenchymal stem cells (BMSCs), derived from TIONFH rabbits, were used for in vitro experiments. Cell viability was determined using the MTT assay. RESULTS: High expression of miR-320a inhibited the osteogenic differentiation capacity of BMSCs in vitro by inhibiting the expression of the osteoblastic differentiation markers ALP and RUNX2. MiR-320a inhibitor-loaded PEG-PLGA-PLL nanoparticles were constructed with a mean loading efficiency of 1.414 ± 0.160%, and a mean encapsulation efficiency of 93.45 ± 1.24%, which released 50% of the loaded miR-320a inhibitor at day 12 and 80% on day 18. Then, inhibitor release entered the plateau. After treatment with the miR-320a inhibitor nanoparticle, the empty lacunae were decreased in the femoral head tissue of TIONFH rabbits, and the osteoblast surface/bone surface (Ob.S/BS), osteoblast number/bone perimeter (Ob.N/B.Pm), bone volume fraction, and bone mineral density increased. Additionally, the expression of osteogenic markers RUNX2 and ALP was significantly elevated in the TIONFH rabbit model. CONCLUSION: The miR-320a inhibitor-loaded PEG-PLGA-PLL nanoparticle sustained drug release system significantly contributed to bone regeneration in the TIONFH rabbit model, which might be a promising strategy for the treatment of TIONFH. BACKGROUND: This study aimed to explore the effect of a nanomaterial-based miR-320a inhibitor sustained release system in trauma-induced osteonecrosis of the femoral head (TIONFH). METHODS: The miR-320a inhibitor-loaded polyethylene glycol (PEG)- Poly(lactic-co-glycolic acid) (PLGA)- Poly-L-lysine (PLL) nanoparticles were constructed using the double emulsion method. The TIONFH rabbit model was established to observe the effects of miR-320a inhibitor nanoparticles in vivo. Hematoxylin–eosin staining and microcomputed tomography scanning were used for bone morphology analysis. Bone marrow mesenchymal stem cells (BMSCs), derived from TIONFH rabbits, were used for in vitro experiments. Cell viability was determined using the MTT assay. RESULTS: High expression of miR-320a inhibited the osteogenic differentiation capacity of BMSCs in vitro by inhibiting the expression of the osteoblastic differentiation markers ALP and RUNX2. MiR-320a inhibitor-loaded PEG-PLGA-PLL nanoparticles were constructed with a mean loading efficiency of 1.414 ± 0.160%, and a mean encapsulation efficiency of 93.45 ± 1.24%, which released 50% of the loaded miR-320a inhibitor at day 12 and 80% on day 18. Then, inhibitor release entered the plateau. After treatment with the miR-320a inhibitor nanoparticle, the empty lacunae were decreased in the femoral head tissue of TIONFH rabbits, and the osteoblast surface/bone surface (Ob.S/BS), osteoblast number/bone perimeter (Ob.N/B.Pm), bone volume fraction, and bone mineral density increased. Additionally, the expression of osteogenic markers RUNX2 and ALP was significantly elevated in the TIONFH rabbit model. CONCLUSION: The miR-320a inhibitor-loaded PEG-PLGA-PLL nanoparticle sustained drug release system significantly contributed to bone regeneration in the TIONFH rabbit model, which might be a promising strategy for the treatment of TIONFH.

On the direct insulator-quantum Hall transition in two-dimensional electron systems in the vicinity of nanoscaled scatterers

Liang, Chi-Te,Lin, Li-Hung,Kuang Yoa, Chen,Lo, Shun-Tsung,Wang, Yi-Ting,Lou, Dong-Sheng,Kim, Gil-Ho,Yuan-Huei, Chang,Ochiai, Yuichi,Aoki, Nobuyuki,Chen, Jeng-Chung,Lin, Yiping,Chun-Feng, Huang,Lin, Sh Springer 2011 Nanoscale research letters Vol.6 No.1

<P>A direct insulator-quantum Hall (I-QH) transition corresponds to a crossover/transition from the insulating regime to a high Landau level filling factor ν > 2 QH state. Such a transition has been attracting a great deal of both experimental and theoretical interests. In this study, we present three different two-dimensional electron systems (2DESs) which are in the vicinity of nanoscaled scatterers. All these three devices exhibit a direct I-QH transition, and the transport properties under different nanaoscaled scatterers are discussed.</P>