Search

The aim of this study is to determine whether the myocardial performance index (MPI) is increased in fetal growth restriction (FGR) fetuses and if increased MPI is related to adverse outcomes of FGR. This is a prospective cross-sectional study. Seventy-three late-onset FGR fetuses and 97 gestational-age matched control fetuses were enrolled in this study. Fetal blood flow parameters including MPI values were measured and compared between the two groups. For the effect of severity of growth restriction on MPI value, they were also compared with < 3rd and 3rd – 10th centile groups. FGR fetuses were divided into two groups by favorable and adverse outcome and ultrasound parameters were compared between these two groups. Moreover, significant factors related to adverse outcomes by univariate analysis were analyzed by multivariate logistic regression analysis. Pulsatility index of umbilical arterial flow (UA-PI), MPI and amniotic fluid index in the FGR were significantly different from the control fetuses. However, no significant difference between < 3rd and 3rd – 10th centile groups was detected in MPI and UA-PI. The increased levels of MPI and UA-PI were independently related with adverse outcome of late-onset FGR pregnancy. In conclusion, MPI values were increased in late-onset FGR pregnancy, and the higher level of MPI could predict adverse outcome as well as the measurement of UA-PI. Clinicians should consider cardiac dysfunction in FGR through increased MPI., This is an Open Access article distributed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view the details of this license, please visit (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Key management is one of the most important tasks in wireless mesh network. This service is responsible for key generation, distribution, and key exchange in a cryptography-based system. Due to the shared nature of WMNs and absence of globally trusted central authority, key management becomes more challenging. This chapter introduces several key management methods that can address these challenges. The fundamental approach is the secret sharing scheme created by A. Shamir, which effectively distributes keys to all participants’ network. Based on Shamir’s scheme, many authors proposed other algorithms to secure the communication channel in such a way that adversary cannot steal any information about the secret. In addition, in this chapter, a new secret sharing method using real-time synchronization among transceiver devices is presented. In this method, each node generates its key depending on its physical information and the real-time clock. Therefore, public and private keys can be managed efficiently for data encryption and prevent several external attacks to WMNs. A specific protocol is proposed to secure keys while transferring between devices to prevent internal attacks.

Local injection of methotrexate (MTX) has been widely used for caesarean scar pregnancy (CSP),but the optimal candidate remains undetermined. The aim of this study is to determine the risk factors associated with treatment failure among patients who received a single dose of local MTX. This is a retrospective cohort study. Clinical information was compared between treatment success vs. failure groups. Risk factors related to treatment failure were also investigated with multivariate analysis. Of 47 patients diagnosed with CSP, 30 received local MTX injection. The initial serum ß- human chorionic gonadotropin (hCG) level in the failure group was significantly higher than in the success group (p = 0.048), and the cut-off value was 47,000 mIU/ml. The rate of type 2 position of the gestational sac in the failure group was significantly higher than in the treatment success group (p = 0.031). A high initial serum ß-hCG level (≥ 47,000 mIU/ml) was identified as the independent risk factor for treatment failure (adjusted odds ratio = 21.9; 95% confidence interval = 1.3–383.1). Type 2 gestational sac position and a higher level of ß-hCG at diagnosis appear to be associated with poor outcomes after local injection of a single dose of MTX., This work was supported by JSPS KAKENHI (Grant Number: 15H02660).

This paper presents a design and implementation of the machinery monitoring and maintenance system. The proposed system has two advanced features including supervision of machinery in real-time and potential fault prediction by using decision tree model. The monitoring system is designed using IoT technologies to collect machine data including power consumption and operation states. Moreover, the potential faults of monitored machines are predicted by the proposed algorithm using decision tree model. The whole system has been implemented and applied in a real factory to demonstrate its performance.

Additional file 1. Data collection tool.

Handwriting recognition is one of the core applications of computer vision for real-word problems and it has been gaining more interest because of the progression in this field. This paper presents an efficient model for Vietnamese handwritten character recognition by Convolutional Neural Networks (CNNs) – a kind of deep neural network model which can achieve high performance on hard recognition tasks. The proposed architecture of the CNN network for Vietnamese handwritten character recognition consists of five hidden layers in which the first 3 layers are convolutional layers and the last 2 layers are fully-connected layers. Overfitting problem is also minimized by using dropout techniques with the reasonable drop rate. The experimental results show that our model achieves approximately 97% accuracy.

Objectives To evaluate the validity of serum CA-125, Human Epididymis protein 4 (HE4) and Risk of Malignancy Algorithm (ROMA) at standard and optimal cut-offs, in preoperative prediction of epithelial ovarian carcinoma (EOC) in Vietnam. Subjects and methods Cross-sectional, descriptive study on 277 patients with ovarian masses hospitalized at the OBGYN Departments, Hue University Hospital and Hue Central Hospital, Vietnam, from 01/2016 to 11/2017. All patients had measurements of serum CA-125 by Elecsys 2010 system and HE4 by immunoassay ARCHITECT® HE4 kits; ROMA calculated, and preoperative malignancy risk estimated. Matching these values to postoperative histopathology resulted in the preoperative prediction values. Results There were 30 (10.8%) cases of EOC. Median values of CA 125, HE4, and ROMA of EOC and benign tumors were 214.20 U/ml, 18.91 U/ml; 90.00 pmol/l, 39.80 pmol/l; and 55.20%, 4.80%, respectively. The sensitivities and specificity of CA125, HE4, and ROMA to distinguish between malignant and benign tumors at standard cut-offs were 83.3% and 78.5%; 50% and 98.38%; 80.0% and 84,6%, and those at optimal cut-offs were 83.3% and 86.6%; 80.0% and 91.5%, 86.7% and 88.7%, respectively. AUCs of CA-125, HE4, and ROMA were 0.872, 0.894, 0.912; and those for the post-menopausal group were 0.900, 0.894 and 0.924, respectively. Conclusion Serum CA 125 and HE4 levels and ROMA have good validity in the diagnosis of EOC, of which ROMA gives the best result. The ROMA index should be applied in clinical practice to help in the assessment and management of patients with suspected ovarian cancer., Highlights • HE4 and ROMA yielded good validity in the preoperative diagnosis of epithelial ovarian cancer. • In pre-menopausal group, the modified ROMA cut-off value (9.89%) yielded a better specificity. • ROMA can help with the assessment and management of patients with suspected ovarian cancer.

In this paper, we present a facial expression recognition (FER) system implemented on a SoC FPGA device. Our system can operate independently and automatically classify 7 different types of basic emotions. We designed a processing engine on FPGA that is able to calculate convolutional layers of convolutional neural networks and designed a CNN optimal with a designed hardware for the task of emotions recognition and achieving 66% of accuracy in FER2013 dataset. The whole hardware system is designed on SoC FPGA which can process up to 15 image frames per second at an operating frequency of 130 MHz.

Advertisements play a crucial role in the mass spreading and success of a product, a service or an entire business. For gaining effectiveness, they should be judged thoroughly by means of direct and indirect rating, portfolio tests, transactions and interactions. Instead of those, human emotions can be a powerful source of reliable feedback. This information is valuable for company to not only re-evaluate the efficiency of their ads but also make them more interactive to customers. To build a good facial expression recognition program, deep learning is applied in the form of Convolutional Neural Network, increasing the accuracy of classifying 7 basic emotions to 90% on CK+ dataset. Moreover, we have designed 2 systems, one for storing and updating the commercials and one for extracting emotion data and attracting customers.

Convolutional neural network (CNN) is one of the most promising algorithms that outweighs other traditional methods in terms of accuracy in classification tasks. However, several CNNs, such as VGG, demand a huge computation in convolutional layers. Many accelerators implemented on powerful FPGAs have been introduced to address the problems. In this paper, we present a VGG-based accelerator which is optimized for a low-cost FPGA. In order to optimize the FPGA resource of logic element and memory, we propose a dedicated input buffer that maximizes the data reuse. In addition, we design a low resource processing engine with the optimal number of Multiply Accumulate (MAC) units. In the experiments, we use VGG16 model for inference to evaluate the performance of our accelerator and achieve a throughput of 38.8[Formula: see text]GOPS at a clock speed of 150[Formula: see text]MHz on Intel Cyclone V SX SoC. The experimental results show that our design is better than previous works in terms of resource efficiency.

This paper presents a new edge-preserving algorithm and its VLSI architecture for block artifact reduction. Unlike previous approaches using block classification, our algorithm utilizes pixel classification to categorize each pixel into one of two classes, namely smooth region and edge region, which are described by the edge-preserving maps. Based on these maps, a two-step adaptive filter which includes offset filtering and edge-preserving filtering is used to remove block artifacts. A pipelined VLSI architecture of the proposed deblocking algorithm for HD video processing is also presented in this paper. A memory-reduced architecture for a block buffer is used to optimize memory usage. The architecture of the proposed deblocking filter is prototyped on FPGA Cyclone II, and then we estimated performance when the filter is synthesized on ANAM 0.25 CMOS cell library using Synopsys Design Compiler. Our experimental results show that our proposed algorithm effectively reduces block artifacts while preserving the details.

This paper presents a new edge-protection algorithm and its very large scale integration (VLSI) architecture for block artifact reduction. Unlike previous approaches using block classification, our algorithm utilizes pixel classification to categorize each pixel into one of two classes, namely smooth region and edge region, which are described by the edge-protection maps. Based on these maps, a two-step adaptive filter which includes offset filtering and edge-preserving filtering is used to remove block artifacts. A pipelined VLSI architecture of the proposed deblocking algorithm for HD video processing is also presented in this paper. A memory-reduced architecture for a block buffer is used to optimize memory usage. The architecture of the proposed deblocking filter is verified on FPGA Cyclone II and implemented using the ANAM 0.25 µm CMOS cell library. Our experimental results show that our proposed algorithm effectively reduces block artifacts while preserving the details. The PSNR performance of our algorithm using pixel classification is better than that of previous algorithms using block classification.