Enhancing the performance of premature ventricular contraction detection in unseen datasets through deep learning with denoise and contrast attention module.

Premature ventricular contraction (PVC) is a common and harmless cardiac arrhythmia that can be asymptomatic or cause palpitations and chest pain in rare instances. However, frequent PVCs can lead to more serious arrhythmias, such as atrial fibrillation. Several PVC detection models have been proposed to enable early diagnosis of arrhythmias; however, they lack reliability and generalizability due to the variability of electrocardiograms across different settings and noise levels. Such weaknesses are known to aggravate with new data. Therefore, we present a deep learning model with a novel attention mechanism that can detect PVC accurately, even on unseen electrocardiograms with various noise levels. Our method, called the Denoise and Contrast Attention Module (DCAM), is a two-step process that denoises signals with a convolutional neural network (CNN) in the frequency domain and attends to differences. It focuses on differences in the morphologies and intervals of the remaining beats, mimicking how trained clinicians identify PVCs. Using three different encoder types, we evaluated 1D U-Net with DCAM on six external test datasets. The results showed that DCAM significantly improved the F1-score of PVC detection performance on all six external datasets and enhanced the performance of balancing both the sensitivity and precision of the models, demonstrating its robustness and generalization ability regardless of the encoder type. This demonstrates the need for a trainable denoising process before applying the attention mechanism. Our DCAM could contribute to the development of a reliable algorithm for cardiac arrhythmia detection under real clinical electrocardiograms.

Stimuli-Responsive, Shape-Transforming Nanostructured Particles.

Development of particles that change shape in response to external stimuli has been a long-thought goal for producing bioinspired, smart materials. Herein, the temperature-driven transformation of the shape and morphology of polymer particles composed of polystyrene-b-poly(4-vinylpyridine) (PS-b-P4VP) block copolymers (BCPs) and temperature-responsive poly(N-isopropylacrylamide) (PNIPAM) surfactants is reported. PNIPAM acts as a temperature-responsive surfactant with two important roles. First, PNIPAM stabilizes oil-in-water droplets as a P4VP-selective surfactant, creating a nearly neutral interface between the PS and P4VP domains together with cetyltrimethylammonium bromide, a PS-selective surfactant, to form anisotropic PS-b-P4VP particles (i.e., convex lenses and ellipsoids). More importantly, the temperature-directed positioning of PNIPAM depending on its solubility determines the overall particle shape. Ellipsoidal particles are produced above the critical temperature, whereas convex lens-shaped particles are obtained below the critical temperature. Interestingly, given that the temperature at which particle shape change occurs depends solely on the lower critical solution temperature (LCST) of the polymer surfactants, facile tuning of the transition temperature is realized by employing other PNIPAM derivatives with different LCSTs. Furthermore, reversible transformations between different shapes of PS-b-P4VP particles are successfully demonstrated using a solvent-adsorption annealing with chloroform, suggesting great promise of these particles for sensing, smart coating, and drug delivery applications.

Morphological Evolution of Block Copolymer Particles: Effect of Solvent Evaporation Rate on Particle Shape and Morphology.

Shape and morphology of polymeric particles are of great importance in controlling their optical properties or self-assembly into unusual superstructures. Confinement of block copolymers (BCPs) in evaporative emulsions affords particles with diverse structures, including prolate ellipsoids, onion-like spheres, oblate ellipsoids, and others. Herein, we report that the evaporation rate of solvent from emulsions encapsulating symmetric polystyrene-b-polybutadiene (PS-b-PB) determines the shape and internal nanostructure of micron-sized BCP particles. A distinct morphological transition from the ellipsoids with striped lamellae to the onion-like spheres was observed with decreasing evaporation rate. Experiments and dissipative particle dynamics (DPD) simulations showed that the evaporation rate affected the organization of BCPs at the particle surface, which determined the final shape and internal nanostructure of the particles. Differences in the solvent diffusion rates in PS and PB at rapid evaporation rates induced alignment of both domains perpendicular to the particle surface, resulting in ellipsoids with axial lamellar stripes. Slower evaporation rates provided sufficient time for BCP organization into onion-like structures with PB as the outermost layer, owing to the preferential interaction of PB with the surroundings. BCP molecular weight was found to influence the critical evaporation rate corresponding to the morphological transition from ellipsoid to onion-like particles, as well as the ellipsoid aspect ratio. DPD simulations produced morphologies similar to those obtained from experiments and thus elucidated the mechanism and driving forces responsible for the evaporation-induced assembly of BCPs into particles with well-defined shapes and morphologies.

Particles with Tunable Porosity and Morphology by Controlling Interfacial Instability in Block Copolymer Emulsions.

A series of porous block copolymer (BCP) particles with controllable morphology and pore sizes was fabricated by tuning the interfacial behavior of BCP droplets in oil-in-water emulsions. A synergistic adsorption of polystyrene-b-poly(4-vinylpyridine) (PS-b-P4VP) BCPs and sodium dodecyl sulfate (SDS) to the surface of the emulsion droplet induced a dramatic decrease in the interfacial tension and generated interfacial instability at the particle surface. In particular, the SDS concentration and the P4VP volume fraction of PS-b-P4VP were key parameters in determining the degree of interfacial instability, leading to different types of particles including micelles, capsules, closed-porosity particles, and open-porosity particles with tunable pore sizes ranging from 10 to 500 nm. The particles with open-porosity could be used as pH-responsive, high capacity delivery systems where the uptake and release of multiple dyes could be achieved.

Size-controlled nanoparticle-guided assembly of block copolymers for convex lens-shaped particles.

The tuning of interfacial properties at selective and desired locations on the particles is of great importance to create the novel structured particles by breaking the symmetry of their surface property. Herein, a dramatic transition of both the external shape and internal morphology of the particles of polystyrene-b-poly(4-vinylpyridine) (PS-b-P4VP) was induced by precise positioning of size-controlled Au nanoparticle surfactants (Au NPs). The size-dependent assembly of the Au NPs was localized preferentially at the interface between the P4VP domain at the particle surface and the surrounding water, which generated a balanced interfacial interaction between two different PS/P4VP domains of the BCP particles and water, producing unique convex lens-shaped BCP particles. In addition, the neutralized interfacial interaction, in combination with the directionality of the solvent-induced ordering of the BCP domains from the interface of the particle/water, generated defect-free, vertically ordered porous channels within the particles. The mechanism for the formation of these novel nanostructures was investigated systemically by varying the size and the volume fraction of the Au NPs. Furthermore, these convex lens-shaped particles with highly ordered channels can be used as a microlens, in which the light can be concentrated toward the focal point with enhanced near-field signals. And, these particles can possess additional optical properties such as unique distribution of light scattering as a result of the well-ordered Au cylinders that filled into the channels, which hold great promise for use in optical, biological-sensing, and imaging applications.