Ensembles of Deep Learning Framework for Stomach Abnormalities Classification

Abnormalities of the gastrointestinal tract are widespread worldwide today. Generally, an effective way to diagnose these life-threatening diseases is based on endoscopy, which comprises a vast number of images. However, the main challenge in this area is that the process is time-consuming and fatiguing for a gastroenterologist to examine every image in the set. Thus, this led to the rise of studies on designing AI-based systems to assist physicians in the diagnosis. In several medical imaging tasks, deep learning methods, especially convolutional neural networks (CNNs), have contributed to the state-of-the-art outcomes, where the complicated nonlinear relation between target classes and data can be learned and not limit to hand-crafted features. On the other hand, hyperparameters are commonly set manually, which may take a long time and leave the risk of non-optimal hyperparameters for classification. An effective tool for tuning optimal hyperparameters of deep CNN is Bayesian optimization. However, due to the complexity of the CNN, the network can be regarded as a black-box model where the information stored within it is hard to interpret. Hence, Explainable Artificial Intelligence (XAI) techniques are applied to overcome this issue by interpreting the decisions of the CNNs in such wise the physicians can trust. To play an essential role in real-time medical diagnosis, CNN-based models need to be accurate and interpretable, while the uncertainty must be handled. Therefore, a novel method comprising of three phases is proposed to classify these life-threatening diseases. At first, hyperparameter tuning is performed using Bayesian optimization for two state-of-the-art deep CNNs, and then Darknet53 and InceptionV3 features are extracted from these fine-tunned models. Secondly, XAI techniques are used to interpret which part of the images CNN takes for feature extraction. At last, the features are fused, and uncertainties are handled by selecting entropy-based features. The experimental results show that the proposed method outperforms existing methods by achieving an accuracy of 97% based on a Bayesian optimized Support Vector Machine classifier.

     

Control charts for the shape parameter of reflected power function distribution under classical estimators

The reflected power function distribution (RPFD) has applications in the fields of reliability engineering and survival analysis. To identify and remove the variation in different reliability processes and also to monitor the reliability of machines where the number of errors follows RPFD, we develop control charts to keep the process in control. A memory less control chart like a Shewhart control chart, and two memory-based control charts like an exponentially weighted moving average (EWMA) control chart and a hybrid exponentially weighted moving average (HEWMA) control chart are discussed and compared with each other. Proposal of these control charts is based on two different estimators, the percentile estimator (PE) and the modified maximum likelihood estimator (MMLE). This study shows that an HEWMA control chart based on PE performs better than PE-based Shewhart and EWMA control charts, as well as MMLE-based Shewhart, EWMA, and HEWMA control charts.     

Improved performance of Bis-GMA dental composites reinforced with surface-modified PAN nanofibers

Journal of Materials Science: Materials in Medicine - Skeletal muscle is an electrically and mechanically active tissue that contains highly oriented, densely packed myofibrils. The tissue has...     

Eu3+, Sm3+ Deep‐Red Phosphors as Novel Materials for White Light‐Emitting Diodes and Simultaneous Performance Enhancement of Organic–Inorganic Perovskite Solar Cells

The luminous efficiency of inorganic white light‐emitting diodes, to be used by the next generation as light initiators, is continuously progressing and is an emerging interest for researchers. However, low color‐rendering index (Ra), high correlated color temperature (CCT), and poor stability limit its wider application. Herein, it is reported that Sm³⁺‐ and Eu³⁺‐doped calcium scandate (CaSc₂O₄ (CSO)) are an emerging deep‐red‐emitting material with promising light absorption, enhanced emission properties, and excellent thermal stability that make it a promising candidate with potential applications in emission display, solid‐state white lighting, and the device performance of perovskite solar cells (PSCs). The average crystal structures of Sm³⁺‐doped CSO are studied by synchrotron X‐ray data that correspond to an extremely rigid host structure. Samarium ion is incorporated as a sensitizer that enhances the emission intensity up to 30%, with a high color purity of 88.9% with a 6% increment. The impacts of hosting the sensitizer are studied by quantifying the lifetime curves. The CaSc₂O₄:0.15Eu³⁺,0.03Sm³⁺ phosphor offers significant resistance to thermal quenching. The incorporation of lanthanide ion‐doped phosphors CSOE into PSCs is investigated along with their potential applications. The CSOE‐coated PSCs devices exhibit a high current density and a high power conversion efficiency (15.96%) when compared to the uncoated control devices.     

Development of AZ91/SiC surface composite by FSP: effect of vibration and process parameters on microstructure and mechanical characteristics

A surface composite layer enhances the mechanical characteristics of a surface while retaining the properties of the base material. Friction stir processing (FSP) is a method for forming surface metal matrix composites (SMMCs) that reinforce a surface with particles. In the current study, a new method entitled friction stir vibration processing (FSVP) was applied to form SMMCs on the surface of AZ91 magnesium alloy with SiC particles as the reinforcing particles. Contrary to FSP, in FSVP, the workpiece was vibrated normal to the processing line while the tool rotated and traversed. The microstructure and mechanical properties of friction stir (FS) and friction stir vibration (FSV) processed specimens were evaluated. Additionally, the effects of vibration frequency and process parameters on different characteristics of FS and FSV processed specimens were studied. The results showed that the stir zone grains for FSV processed specimens were finer than those for FS processed specimens, and the second phase particles (SiC particles) had a more homogenous distribution in the former specimens than in the latter specimens. This was related to the effect of workpiece vibration during FSVP, which increased the material deformation and led to enhanced dynamic recrystallization and the breakdown of agglomerated SiC particles. The results indicated that the stir zone grain size decreased, and the distribution homogeneity of the SiC particles increased as vibration frequency increased. It was also observed that the stir zone grain size increased, and the mechanical properties of the processed specimens decreased as tool rotation speed increased.The full text can be downloaded at https://link.springer.com/content/pdf/10.1007%2Fs40436-019-00288-9.pdf     

Maximal atom–photon entanglement in a double-$$\Lambda $$ quantum system

The atom–photon entanglement of a dressed atom and its spontaneous emission in a double-\(\Lambda \) closed-loop atomic system is studied under multi-photon resonance condition. It is shown that even in the absence of quantum interference due to the spontaneous emission, the von Neumann entropy is phase-sensitive and it can be controlled by either intensity or relative phase of the applied fields. It is demonstrated that for the special case of Rabi frequency of the applied fields, the system is maximally entangled. Moreover, an open-loop configuration is considered, and it is shown that the degree of entanglement can be controlled by intensity of the applied fields. Furthermore, in electromagnetically induced transparency condition, the system is disentangled. Such a system can be used for quantum information processing via entanglement using optical switching.     

Introducing a Novel Hybrid Machine Learning Model and Developing its Performance in Estimating Water Quality Parameters

Water Resources Management - For the first time, a novel hybrid machine learning model named the least-squares support vector machine-arithmetic optimization algorithm (LSSVM-AOA) was proposed. The...     

Mixed oxide nanotubes in nanomedicine: A dead-end or a bridge to the future?

Nanomedicine has seen a significant rise in the development of new research tools and clinically functional devices. In this regard, significant advances and new commercial applications are expected in the pharmaceutical and orthopedic industries. For advanced orthopedic implant technologies, appropriate nanoscale surface modifications are highly effective strategies and are widely studied in the literature for improving implant performance. It is well-established that implants with nanotubular surfaces show a drastic improvement in new bone creation and gene expression compared to implants without nanotopography. Nevertheless, the scientific and clinical understanding of mixed oxide nanotubes (MONs) and their potential applications, especially in biomedical applications are still in the early stages of development. This review aims to establish a credible platform for the current and future roles of MONs in nanomedicine, particularly in advanced orthopedic implants. We first introduce the concept of MONs and then discuss the preparation strategies. This is followed by a review of the recent advancement of MONs in biomedical applications, including mineralization abilities, biocompatibility, antibacterial activity, cell culture, and animal testing, as well as clinical possibilities. To conclude, we propose that the combination of nanotubular surface modification with incorporating sensor allows clinicians to precisely record patient data as a critical contributor to evidence-based medicine.     

Memristive fuzzy edge detector

Fuzzy inference systems always suffer from the lack of efficient structures or platforms for their hardware implementation. In this paper, we tried to overcome this difficulty by proposing a new method for the implementation of the fuzzy rule-based inference systems. To achieve this goal, we have designed a multi-layer neuro-fuzzy computing system based on the memristor crossbar structure by introducing a new concept called the fuzzy minterm. Although many applications can be realized through the use of our proposed system, in this study we only show how the fuzzy XOR function can be constructed and how it can be used to extract edges from grayscale images. One main advantage of our memristive fuzzy edge detector (implemented in analog form) compared to other commonly used edge detectors is it can be implemented in parallel form, which makes it a powerful device for real-time applications.     

Spatially homogeneous gelation in liquid composite molding

On‐line mixing of the resin with its curing agents prior to injection into a mold is a common industrial technique for fabricating composite parts. For vinyl‐ester resins that cure via free radical polymerization, the concentrations of retarder, accelerator, and initiator are pre‐selected and cannot be changed during the injection. Hence, the resin that enters the mold the earliest has cured longer than the resin that enters the mold later, since the gel time for the resin is the same, owing to the fixed ratio of the curing agents. This approach leads to inhomogeneous cure of the resin and consequently to longer residence time of the resin in the mold. It requires an additional 50 to 75 percent of the filling time before a part can be de‐molded. In this study, it is shown that by adjusting the concentration of curing agents during the injection, a more homogeneous gel time throughout the mold can be achieved. The time to de‐mold is reduced to 18‐24 percent of the filling time. Sensors that measure the conductivity of the resin were used to detect the location and monitor the cure of vinyl‐ester. This approach could be extended to other resin systems to control the spatial curing of the resin in the mold.