Capsule Architectures Research Articles

Deep learning modelling techniques: current progress, applications, advantages, and challenges

Deep learning (DL) is revolutionizing evidence-based decision-making techniques that can be applied across various sectors. Specifically, it possesses the ability to utilize two or more levels of non-linear feature transformation of the given data via representation learning in order to overcome limitations posed by large datasets. As a multidisciplinary field that is still in its nascent phase, articles that survey DL architectures encompassing the full scope of the field are rather limited. Thus, this paper comprehensively reviews the state-of-art DL modelling techniques and provides insights into their advantages and challenges. It was found that many of the models exhibit a highly domain-specific efficiency and could be trained by two or more methods. However, training DL models can be very time-consuming, expensive, and requires huge samples for better accuracy. Since DL is also susceptible to deception and misclassification and tends to get stuck on local minima, improved optimization of parameters is required to create more robust models. Regardless, DL has already been leading to groundbreaking results in the healthcare, education, security, commercial, industrial, as well as government sectors. Some models, like the convolutional neural network (CNN), generative adversarial networks (GAN), recurrent neural network (RNN), recursive neural networks, and autoencoders, are frequently used, while the potential of other models remains widely unexplored. Pertinently, hybrid conventional DL architectures have the capacity to overcome the challenges experienced by conventional models. Considering that capsule architectures may dominate future DL models, this work aimed to compile information for stakeholders involved in the development and use of DL models in the contemporary world.

Microfluidic Fabrication of Phase-Inverted Microcapsules with Asymmetric Shell Membranes with Graded Porosity.

Microcapsules with liquid cores and solid shells are attractive as dispersible protective micron-sized containers. Applications that rely on molecular mass transport often require a combination of size selectivity, high permeability, and mechanical stability. Capsule architectures that combine all these features represent a material property, design, and fabrication challenge. In this work, the design of an asymmetric microcapsule shell architecture is reported to achieve a good combination of the desired features. Poly(methyl methacrylate) phase-inverted microcapsules featuring an asymmetric graded macroporous shell covered with a dense skin separation layer are obtained from water-in-oil-in-water double emulsion drops that are phase-inverted in a water-based coagulation bath. The phase-inverted microcapsules exhibit good mechanical stability and allow for high permeability of its shell membrane with molecular size dependence.

An EM Algorithm for Capsule Regression.

We investigate a latent variable model for multinomial classification inspired by recent capsule architectures for visual object recognition (Sabour, Frosst, & Hinton, 2017). Capsule architectures use vectors of hidden unit activities to encode the pose of visual objects in an image, and they use the lengths of these vectors to encode the probabilities that objects are present. Probabilities from different capsules can also be propagated through deep multilayer networks to model the part-whole relationships of more complex objects. Notwithstanding the promise of these networks, there still remains much to understand about capsules as primitive computing elements in their own right. In this letter, we study the problem of capsule regression-a higher-dimensional analog of logistic, probit, and softmax regression in which class probabilities are derived from vectors of competing magnitude. To start, we propose a simple capsule architecture for multinomial classification: the architecture has one capsule per class, and each capsule uses a weight matrix to compute the vector of hidden unit activities for patterns it seeks to recognize. Next, we show how to model these hidden unit activities as latent variables, and we use a squashing nonlinearity to convert their magnitudes as vectors into normalized probabilities for multinomial classification. When different capsules compete to recognize the same pattern, the squashing nonlinearity induces nongaussian terms in the posterior distribution over their latent variables. Nevertheless, we show that exact inference remains tractable and use an expectation-maximization procedure to derive least-squares updates for each capsule's weight matrix. We also present experimental results to demonstrate how these ideas work in practice.

Coherent X-ray diffraction imaging of nanoengineered polymeric capsules

For the first time, nanoengineered polymeric capsules and their architecture have been studied with coherent X-ray diffraction imaging technique. The use of coherent X-ray diffraction imaging technique allowed us to analyze the samples immersed in a liquid. We report about the significant difference between polymeric capsule architectures under dry and liquid conditions.

Stimuli-Responsive Capsule Membranes for Controlled Release in Pharmaceutical Applications.

In conventional drug delivery, the drug concentration in the blood raises once the drug taken, and then peaks and declines. Since each drug has a level above which it is toxic and another level below which it is ineffective, the drug concentration in a patient at a particular time depends on compliance with the prescribed routine. To achieve more effective efficacy and fewer side effects of drugs, the drug carriers with desirable dosing and controllable release property of drugs are highly desired. Stimuli-responsive capsules with smart gating membranes or hydrogel-based membranes as capsule shells are ideal candidates. The smart capsule membranes enable efficient encapsulation of drugs within the large inner volume, and the responsive gating membranes or hydrogel-based membranes could control the release rate of encapsulated drugs in responding to environmental stimuli. The trigger stimuli could be either artificial or natural ones corresponding to specific diseases, such as temperature, pH, glucose concentration, specific ion, light, and magnetic field. This review highlights the recent development in stimuli-responsive capsule membranes for controlled release in pharmaceutical applications, including two types of stimuli-responsive capsule membranes with different architectures for on/off release and burst release, which can achieve potential uses of case-dependent on/off release and burst release. The preponderances of the smart capsule membranes are that the capsules are with controllable inner space for drug vehicles with desired dose and stimuli-responsive membrane as shell to release drugs at a desired site and/or moment. However, the actual difficulties for the stimuli-responsive capsule membrane systems to go before they can be applied widely in the biomedical fields are discussed. The future works should focus on the improvements of biocompatibility, biodegradability and stimuli-responsiveness of the capsule membranes, easy and scalable fabrication techniques with further decrease of the capsule size for more efficient in vivo applications, and the diversification of the multi-compartmental capsule architectures with multi-stimuli-responsive characteristics for controlled release.

Crystallization-Driven Multicomponent Encapsulation of Coulombically Repulsive Guests

Coulombically repulsive cationic guests can be encapsulated in the multicomponent capsule hosts via favorable ion pairing, H-bonding intermolecular interactions. The inclusion of protonated and diprotonated 2-pycolyl amine guests in the calix[4]arene tetrasulfonate, tetrasodium salt C[4]) host is reminiscent with complex dynamic equilibria in solution. The addition of AG aminoguanidinium cations to such mixtures results in the formation of distinct crystalline bilayer and capsule architectures in the solid state. This show that libraries of interexchanging supramolecular components may convert into robust solid state multicomponent capsules via constitutional crystallization.