Composing RESTful Services and Collaborative Workflows: A Lightweight Approach

The use of RESTful Web services has gained momentum in the development of distributed applications based on traditional Web standards such as HTTP. In particular, these services can integrate easily into various applications, such as mashups. Composing RESTful services into Web-scale workflows requires a lightweight composition language that's capable of describing both the control and data flow that constitute a workflow. The authors address these issues with Bite, a lightweight and extensible composition language that enables the creation of Web-scale workflows and uses RESTful services as its main composable entities.     

Visual tracking using convolutional features with sparse coding

Artificial Intelligence Review - Visual object tracking has become one of the most active research topics in computer vision, and it has been applied in several commercial...     

Optimal Artificial Intelligence Based Automated Skin Lesion Detection and Classification Model

Skin lesions have become a critical illness worldwide, and the earlier identification of skin lesions using dermoscopic images can raise the survival rate. Classification of the skin lesion from those dermoscopic images will be a tedious task. The accuracy of the classification of skin lesions is improved by the use of deep learning models. Recently, convolutional neural networks (CNN) have been established in this domain, and their techniques are extremely established for feature extraction, leading to enhanced classification. With this motivation, this study focuses on the design of artificial intelligence (AI) based solutions, particularly deep learning (DL) algorithms, to distinguish malignant skin lesions from benign lesions in dermoscopic images. This study presents an automated skin lesion detection and classification technique utilizing optimized stacked sparse autoencoder (OSSAE) based feature extractor with backpropagation neural network (BPNN), named the OSSAE-BPNN technique. The proposed technique contains a multi-level thresholding based segmentation technique for detecting the affected lesion region. In addition, the OSSAE based feature extractor and BPNN based classifier are employed for skin lesion diagnosis. Moreover, the parameter tuning of the SSAE model is carried out by the use of sea gull optimization (SGO) algorithm. To showcase the enhanced outcomes of the OSSAE-BPNN model, a comprehensive experimental analysis is performed on the benchmark dataset. The experimental findings demonstrated that the OSSAE-BPNN approach outperformed other current strategies in terms of several assessment metrics.     

Proposed network structures and combined adaptive algorithms for electrocardiogram signal denoising

An electrocardiogram (ECG) signal is a record of the electrical activities of heart muscle and is used clinically to diagnose heart diseases. An ECG signal should be presented as clear as possible to support accurate decisions made by doctors. This article proposes different combinations of combined adaptive algorithms to derive different noise-cancelling structures to remove (denoise) different kinds of noise from ECG signals. The algorithms are applied to the following types of noise: power line interference, baseline wander, electrode motion artifact, and muscle artifacts. Moreover, the results of the suggested models and algorithms are compared with those of conventional denoising tools such as the discrete wavelet transform, an adaptive filter, and a multilayer neural network (NN) to ensure the superiority of the proposed combined structures and algorithms. Furthermore, the hybrid concept is based on dual, triple, and quadruple combinations of well-known algorithms that derive adaptive filters, such as the least mean squares, normalized least mean squares and recursive least squares algorithms. The combinations are formulated based on partial update, variable step-size (VSS), and second iterative VSS algorithms, which are considered in different combinations. In addition, biased NN and unbiased linear neural network (ULNN) structures are considered. The performance of the different structures and related algorithms are evaluated by measuring the post-signal-to-noise ratio, mean square error, and percentage root mean square difference.     

Improving PAPR performance of filtered OFDM for 5G communications using PTS

The filtered orthogonal frequency division multiplexing (F-OFDM) system has been recommended as a waveform candidate for fifth-generation (5G) communications. The suppression of out-of-band emission (OOBE) and asynchronous transmission are the distinctive features of the filtering-based waveform frameworks. Meanwhile, the high peak-to-average power ratio (PAPR) is still a challenge for the new waveform candidates. Partial transmit sequence (PTS) is an effective technique for mitigating the trend of high PAPR in multicarrier systems. In this study, the PTS technique is employed to reduce the high PAPR value of an F-OFDM system. Then, this system is compared with the OFDM system. In addition, the other related parameters such as frequency localization, bit error rate (BER), and computational complexity are evaluated and analyzed for both systems with and without PTS. The simulation results indicate that the F-OFDM based on PTS achieves higher levels of PAPR, BER, and OOBE performances compared with OFDM. Moreover, the BER performance of F-OFDM is uninfluenced by the use of the PTS technique.     

An experimental study for the selection of modules and facilities in a mass customization context

To design an efficient product family, designers have to anticipate the production process and, more generally, the supply chain costs. But this is a difficult problem, and designers often propose a solution which is subsequently evaluated in terms of logistical costs. This paper presents a design problem in which the product and the supply chain design are considered at the same time. It consists in selecting a set of modules that will be manufactured at distant facilities and then shipped to a plant close to the market for final, customized assembly under time constraints. The goal is to obtain the bill of materials for all the items in the product family, each of which is made up of a set of modules, and specifying the location where these modules will be built, in order to minimize the total production costs for the supply chain. The objective of the study is to analyze both, for small instances, the impact of the costs (fixed and variable) on the optimal solutions, and to compare an integrated approach minimizing the total cost in one model with a two-phases approach in which the decisions relating to the design of the products and the allocation of modules to distant sites are made separately.     

A real-time industrial pattern classification system

Using the partitioned matrix approach, a parallel hardware architecture for a parametric (Bayes) classifier is designed. The architecture consists of simple, regularly structured processing elements operating in parallel. As a result, the proposed design is suitable for VLSI implementation. A comparative analysis shows that the approach is more efficient and can significantly reduce the cost required for implementing the classifier, while maintaining high speed     

A new software for initiating and optimising insulin treatment of out-patients

Two computer programs were developed, program I to optimise insulin treatment using six injections per day, and program II to convert these insulin profiles into less frequent injections of mixtures of regular and NPH insulin. The first software in an HP 41 CV pocket computer uses iterative adjustments during the day and on subsequent days to determine the optimal timing and dosage of insulin. Six self-monitored glucose values at 3 h intervals, insulin doses, and the effects of insulin on plasma glucose are memorised for calculations. The calculated insulin doses were applied by Optipen as five s.c. injections of regular insulin and one bedside injection of NPH insulin. After 5 days the optimised individual insulin profiles with six daily injections were processed by program II. It applies the pharmacokinetics of regular and NPH insulin to make suggestions for a more conventional insulin therapy with one, two, three or four daily injections of regular insulin, NPH insulin, or varying mixtures of both insulins. The procedure was well tolerated in eight insulin-dependent diabetes mellitus out-patients. Insulin therapy with two or three injections, fitted by the second program and selected according to the quality score, produced plasma glucose profiles as satisfactory as those obtained with six injections. The system allows the fine tuning of insulin therapy to out-patients with their individual diets and physical activities.     

Factors affecting the outcome of foetal hydrocephaly

In this study, the authors attempt to provide an account of the factors that affect the outcome of hydrocephaly in 26 foetuses. The hydrocephalus was related to a myelomeningocele in 35% of patients. Sixty-two percent of foetuses showed intra-uterine progression of their hydrocephalus and 50% were shunted postnatally. At a mean follow up of 2 years, the outcome was considered "fair" in 54% of patients. Our findings demonstrate that in foetal hydrocephaly a more favourable outcome is expected in patients with hydrocephalus which does not progress in utero, in whom the labour is not induced before 36 weeks of gestation, who are delivered vaginally weighing more than 2.5 kg and have a head circumference below the 95th centile and a CT cortical mantle thickness of 2 cm and more and who are treated by CSF shunting. The diagnosis of the foetal hydrocephaly at or before 28 weeks of gestation and the presence of a myelomeningocele did not affect the outcome significantly. Neurosurgeons are reminded to keep an open mind for infants with foetal hydrocephaly and to offer active treatment to patients with a potentially favourable outcome.