Flow-focusing microfluidic devices have attracted increasing attention for their applications in droplet generation, used in high-throughput single-cell analysis and mimicking cellular and organelle environments. In addition, there is growing interest in replacing traditional polydimethylsiloxane (PDMS) soft lithography with additive manufacturing, particularly fused filament fabrication (FFF), because of its rapid prototyping capability and low-cost processing. The choice of printing material also affects the scalability and performance of microfluidic devices. Among thermoplastic materials, cyclic olefin copolymer (COC) is a promising alternative because of its high optical clarity, chemical resistance, low water absorption, and biocompatibility. However, fabrication of complex microfluidic junctions remains challenging because conventional slicing strategies can introduce filament artifacts, leakage, and defects at channel intersections. In this work, a four-way flow-focusing microfluidic device was developed. The design and fabrication process was based on a CAD-to-toolpath methodology in which microfluidic features were treated independently during slicing rather than relying solely on conventional global printing parameters. Each channel subcomponent, including the channel window, main channel, and channel roof, was refined with different lengths at each layer to prevent overlap between the branch and straight channels. Slicer settings and post-processing of G-code were used to control the sequence and continuity of extrusion paths in critical channel regions, particularly at the four-way junction. Specifically, the channel subcomponents were printed before the infill regions to maintain the desired channel dimensions and improve sealing against the supporting substrate. The toolpath was further modified to extend the branch roof across the stem roof, maintaining continuity at the channel intersection. In addition, the travel paths at the beginning of the channel walls were adjusted to move out of the main channel, minimizing channel deformation caused by excess material when extrusion resumed. Retraction commands before and after printing the channel components were also adjusted to approximately 0.05 mm greater than the corresponding extrusion commands to minimize overextrusion and underextrusion when extrusion resumed. The resulting device was successfully fabricated with leak-free microfluidic channels. The device was subsequently evaluated for droplet generation by introducing oil and water phases through the flow channels, successfully producing oil-in-water droplets.

Currently, there is no way to quantitively measure bias so the purpose of this was to attempt to link Alpha Amylase 1 (AMY1) to implicit bias. The estimated result states that the study will show an increase in AMY1 levels when presenting stimuli of a man who is black as opposed to AMY1 levels decreasing when presenting stimuli of a man who is white. Participants completed a survey before asking them to self-score their stress level and their opinion stating if they are biased. Then samples were taken before and after being exposed to a video of a situation involving a man who was either white or Black. Results show that the majority of participants experienced a decreased AMY1 level when exposed to a stimulus of a white man signifying decreased stress levels, with an increase in AMY1 levels when exposed to a stimulus of a Black man signifying increased stress levels. Additionally, it was seen that participants preexisting stress level did not impact their results from the study.

Intracranial aneurysms (IAs) with bleb(s)/daughter sacs have higher potential of rupture. Hemodynamics plays a key role in the formation and rupture of aneurysms and daughter sacs in the human vascular system. This investigation studies the hemodynamic factors linked to the initiation of daughter sacs and rupture of IAs with daughter sacs using anatomical and phantom models with the sac virtually removed across various scenarios. Anatomical models of 45 IAs with 67 bleb scenarios were created through reconstructions from 3D rotational angiographies. Corresponding phantom parental IA models, assuming as the pre-bleb state, were created by virtually removing the bleb. Temporal hemodynamic flow features in both blebs and parental IAs under physiological pulsatile inflow conditions were revealed using an in-vitro validated computational fluid dynamics (CFD) model. Statistical analysis were conducted on several hemodynamic factors, such as the wall shear stress (WSS), WSS gradient, time-averaged WSS, surface-averaged WSS, oscillatory shear index etc. It is found that higher WSS was linked to IA rupture at anterior cerebral artery (ACA) bifurcation (p < 0.05), while relatively low WSS was associated with IA rupture on the internal carotid artery (ICA) (p < 0.05). High surface-averaged WSS (p< 0.05) and WSS gradient (p< 0.05) with lower maximum OSI (p< 0.05), were found to significantly increase the potential of daughter sac initiation. The daughter sac is inferred to be initiated near the edge region of the original impacted region where higher WSS was observed. The hemodynamic parameters influencing IA rupture with bleb(s) vary significantly with specific aneurysm locations.

In vitro modeling of barrier tissues like the skin, lungs, gut, and blood-brain barrier using Organ-on-a-Chip (OOC) devices is increasingly important for screening and testing new drugs for efficacy and transport. Current models rely on flat synthetic polymer membranes that fail to accurately replicate the complex, fibrous architecture of the natural human extracellular matrix (ECM). While electrospun nanofiber (ESNF) membranes offer high porosity and controllable structures ideal for mimicking the human ECM, integrating these fragile mats into dielectric devices currently requires manual peeling, cutting, and transferring. This manual handling causes tearing, wrinkling, and poor sealing, ultimately compromising barrier integrity and resulting in inconsistent device fabrication. Consequently, there is a significant need for a process that focuses electrospun nanofiber mats into dielectric OOC devices. To address these critical limitations, we developed an automated, seamlessly integrated manufacturing process using a novel Gel Gap Electrospinning (GGES) technique. In this approach, a 3D printer first fabricates a dielectric polymer device base and a conductive gel-polymer-electrolyte (GPE) targeting ring. The device is then transferred to an electrospinning setup, where the temporary GPE ring creates a localized conductive region that enables nanofibers to be deposited within the gel’s perimeter on the dielectric substrate. Finally, after the GPE is dried, the device is returned to the 3D printer where fabrication is completed by printing the upper half of the device. Here, we report a low-residue GPE formulation engineered to provide suitable rheological and electrical properties and to dry with minimal residue following electrospinning. This minimizes interference with the subsequent 3D printing of the upper Transwell components and facilitates interlayer fusion. Through systematic material optimization, glycerol was replaced with sodium citrate to achieve a viscosity suitable for horizontal electrospinning, helping the hydrogel remain securely in place and reducing macroscopic printing defects such as gaps and bubbles. Furthermore, the incorporation of NaCl increased the formulation’s electrical conductivity 3.54-fold, from 2.5 S/m to 8.85 S/m. Together, these material and process improvements enabled the fabrication of Transwell architectures with targeted nanofiber deposition. Ultimately, the GGES technique provides a practical and potentially scalable manufacturing platform for incorporating fibrous scaffolds into dielectric devices for in vitro screening.

Awaiting public release.

Understanding the thermal properties of porous/multiphase materials is essential in engineering. However, conventional laboratory and in-situ testing methods are often time consuming, labor-intensive, and inadequate for capturing complex multiscale characteristics. To address this gap, this study proposes an innovative, AI-powered framework for the rapid, cost effective, and image-based property evaluation of materials. The framework integrates four components: (1) a novel Voronoi Imaging Method, (2) random finite element modeling (rFEM)-based virtual experiments, (3) advanced machine learning (ML) techniques, and (4) RGB camera imaging. In the presentation, Unsaturated soil was used as a demonstration case, focusing on thermal conductivity estimation. Within the framework, an innovative polygon-filling Voronoi algorithm was proposed to generate virtual 2D image samples of porous materials, accurately replicating their multiscale microstructures. These virtual samples are then used for rFEM virtual experiments to test material properties and generate data for training a one-hot encoded convolutional neural network (CNN). To support image-based property prediction, an RGB–XY Decision Tree Classifier was proposed and used to process RGB photos of real samples to extract 2D phase-distribution maps of air, water, and solid. These maps are then used as input to the trained CNN model, enabling photo-driven estimation of material properties. Results show that the CNN model can predict values closely matching virtual experiment and laboratory results (errors within ±15%).

Soft wearable strain sensors can support continuous joint-motion monitoring for rehabilitation and other wearable applications, but maintaining signal quality during motion and ensuring mechanical durability under repeated deformation remain major challenges. To address these challenges, this study examined whether kirigami cuts could reduce permanent deformation and stabilize the electrical response of eutectic gallium-indium (EGaIn) liquid-metal sensors. Cut and uncut sensors were fabricated using thermoplastic polyurethane (TPU) and evaluated as TPU-only and fabric-backed designs under repeated stretching. Compared with uncut designs, kirigami-patterned sensors withstood up to 660% strain without failure and returned closer to their initial shape after unloading, although they produced a weaker voltage response. Adding fabric backing increased the response of the cut sensors and improved their consistency across cycles. These results suggest that combining kirigami geometry with fabric support can improve sensor reliability while maintaining the measurable electrical response needed for wearable sensing.