Imbalanced spectral data analysis using data augmentation based on the generative adversarial network.

Spectroscopic techniques generate one-dimensional spectra with distinct peaks and specific widths in the frequency domain. These features act as unique identities for material characteristics. Deep neural networks (DNNs) has recently been considered a powerful tool for automatically categorizing experimental spectra data by supervised classification to evaluate material characteristics. However, most existing work assumes balanced spectral data among various classes in the training data, contrary to actual experiments, where the spectral data is usually imbalanced. The imbalanced training data deteriorates the supervised classification performance, hindering understanding of the phase behavior, specifically, sol-gel transition (gelation) of soft materials and glycomaterials. To address this issue, this paper applies a novel data augmentation method based on a generative adversarial network (GAN) proposed by the authors in their prior work. To demonstrate the effectiveness of the proposed method, the actual imbalanced spectral data from Pluronic F-127 hydrogel and Alpha-Cyclodextrin hydrogel are used to classify the phases of data. Specifically, our approach improves 8.8%, 6.4%, and 6.2% of the performance of the existing data augmentation methods regarding the classifier's F-score, Precision, and Recall on average, respectively. Specifically, our method consists of three DNNs: the generator, discriminator, and classifier. The method generates samples that are not only authentic but emphasize the differentiation between material characteristics to provide balanced training data, improving the classification results. Based on these validated results, we expect the method's broader applications in addressing imbalanced measurement data across diverse domains in materials science and chemical engineering.

Synthesis and real-time characterization of self-healing, injectable, fast-gelling hydrogels based on alginate multi-reducing end polysaccharides (MREPs).

Polysaccharide-based hydrogels are promising for many biomedical applications including drug delivery, wound healing, and tissue engineering. We illustrate herein self-healing, injectable, fast-gelling hydrogels prepared from multi-reducing end polysaccharides, recently introduced by the Edgar group. Simple condensation of reducing ends from multi-reducing end alginate (M-Alg) with amines from polyethylene imine (PEI) in water affords a dynamic, hydrophilic polysaccharide network. Trace amounts of acetic acid can accelerate the gelation time from hours to seconds. The fast-gelation behavior is driven by rapid Schiff base formation and strong ionic interactions induced by acetic acid. A cantilever rheometer enables real-time monitoring of changes in viscoelastic properties during hydrogel formation. The reversible nature of these crosslinks (imine bonds, ionic interactions) provides a hydrogel with low toxicity in cell studies as well as self-healing and injectable properties. Therefore, the self-healing, injectable, and fast-gelling M-Alg/PEI hydrogel holds substantial promise for biomedical, agricultural, controlled release, and other applications.

Improving biosensor accuracy and speed using dynamic signal change and theory-guided deep learning.

False results and time delay are longstanding challenges in biosensing. While classification models and deep learning may provide new opportunities for improving biosensor performance, such as measurement confidence and speed, it remains a challenge to ensure that predictions are explainable and consistent with domain knowledge. Here, we show that consistency of deep learning classification model predictions with domain knowledge in biosensing can be achieved by cost function supervision and enables rapid and accurate biosensing using the biosensor dynamic response. The impact and utility of the methodology were validated by rapid and accurate quantification of microRNA (let-7a) across the nanomolar (nM) to femtomolar (fM) concentration range using the dynamic response of cantilever biosensors. Data augmentation and cost function supervision based on the consistency of model predictions and experimental observations with the theory of surface-based biosensors improved the F1 score, precision, and recall of a recurrent neural network (RNN) classifier by an average of 13.8%. The theory-guided RNN (TGRNN) classifier enabled quantification of target analyte concentration and false results with an average prediction accuracy, precision, and recall of 98.5% using the initial transient or entire dynamic response, which is indicative of high prediction accuracy and low probability of false-negative and false-positive results. Classification scores were used to establish new relationships among biosensor performance characteristics (e.g., measurement confidence) and design parameters (e.g., inputs and hyperparameters of classification models and data acquisition parameters) that may be used for characterizing biosensor performance.

Reduction of Biosensor False Responses and Time Delay Using Dynamic Response and Theory-Guided Machine Learning.

Here, we provide a new methodology for reducing false results and time delay of biosensors, which are barriers to industrial, healthcare, military, and consumer applications. We show that integrating machine learning with domain knowledge in biosensing can complement and improve the biosensor accuracy and speed relative to the performance achieved by traditional regression analysis of a standard curve based on the biosensor steady-state response. The methodology was validated by rapid and accurate quantification of microRNA across the nanomolar to femtomolar range using the dynamic response of cantilever biosensors. Theory-guided feature engineering improved the performance and efficiency of several classification models relative to the performance achieved using traditional feature engineering methods (TSFRESH). In addition to the entire dynamic response, the technique enabled rapid and accurate quantification of the target analyte concentration and false-positive and false-negative results using the initial transient response, thereby reducing the required data acquisition time (i.e., time delay). We show that model explainability can be achieved by combining theory-guided feature engineering and feature importance analysis. The performance of multiple classifiers using both TSFRESH- and theory-based features from the biosensor's initial transient response was similar to that achieved using the entire dynamic response with data augmentation. We also show that the methodology can guide design of experiments for high-performance biosensing applications, specifically, the selection of data acquisition parameters (e.g., time) based on potential application-dependent performance thresholds. This work provides an example of the opportunities for improving biosensor performance, such as reducing biosensor false results and time delay, using explainable machine learning models supervised by domain knowledge in biosensing.

Accelerated Engineering of Optimized Functional Composite Hydrogels via High-Throughput Experimentation.

The Materials Genome Initiative (MGI) seeks to accelerate the discovery and engineering of advanced materials via high-throughput experimentation (HTE), which is a challenging task, given the common trade-off between design for optimal processability vs performance. Here, we report a HTE method based on automated formulation, synthesis, and multiproperty characterization of bulk soft materials in well plate formats that enables accelerated engineering of functional composite hydrogels with optimized properties for processability and performance. The method facilitates rapid high-throughput screening of hydrogel composition-property relations for multiple properties in well plate formats. The feasibility and utility of the method were demonstrated by application to several functional composite hydrogel systems, including alginate/poly(N-isopropylacrylamide) (PNIPAM) and poly(ethylene glycol) dimethacrylate (PEGDMA)/poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) hydrogels. The HTE method was leveraged to identify formulations of conductive PEGDMA/PEDOT:PSS composite hydrogels for optimized performance and processability in three-dimensional (3D) printing. This work provides an advance in experimental methods based on automated dispensing, mixing, and sensing for the accelerated engineering of soft functional materials.

A sweet solution to complex microprinting.

A simple sugar mixture transfers functional components to surfaces with intricate geometry.

Macrophage Activation in the Dorsal Root Ganglion in Rats Developing Autotomy after Peripheral Nerve Injury.

Autotomy, self-mutilation of a denervated limb, is common in animals after peripheral nerve injury (PNI) and is a reliable proxy for neuropathic pain in humans. Understanding the occurrence and treatment of autotomy remains challenging. The objective of this study was to investigate the occurrence of autotomy in nude and Wistar rats and evaluate the differences in macrophage activation and fiber sensitization contributing to the understanding of autotomy behavior. Autotomy in nude and Wistar rats was observed and evaluated 6 and 12 weeks after sciatic nerve repair surgery. The numbers of macrophages and the types of neurons in the dorsal root ganglion (DRG) between the two groups were compared by immunofluorescence studies. Immunostaining of T cells in the DRG was also assessed. Nude rats engaged in autotomy with less frequency than Wistar rats. Autotomy symptoms were also relatively less severe in nude rats. Immunofluorescence studies revealed increased macrophage accumulation and activation in the DRG of Wistar rats. The percentage of NF200+ neurons was higher at 6 and 12 weeks in Wistar rats compared to nude rats, but the percentage of CGRP+ neurons did not differ between two groups. Additionally, macrophages were concentrated around NF200-labeled A fibers. At 6 and 12 weeks following PNI, CD4+ T cells were not found in the DRG of the two groups. The accumulation and activation of macrophages in the DRG may account for the increased frequency and severity of autotomy in Wistar rats. Our results also suggest that A fiber neurons in the DRG play an important role in autotomy.

Closed-Loop Controlled Photopolymerization of Hydrogels.

Here, we present a closed-loop controlled photopolymerization process for fabrication of hydrogels with controlled storage moduli. Hydrogel crosslinking was associated with a significant change in the phase angle of a piezoelectric cantilever sensor and established the timescale of the photopolymerization process. The composition, structure, and mechanical properties of the fabricated hydrogels were characterized using Raman spectroscopy, scanning electron microscopy (SEM), and dynamic mechanical analysis (DMA). We found that the storage moduli of photocured poly(ethylene glycol) dimethacrylate (PEGDMA) and poly(N-isopropylacrylamide) (PNIPAm) hydrogels could be controlled using bang-bang and fuzzy logic controllers. Bang-bang controlled photopolymerization resulted in constant overshoot of the storage modulus setpoint for PEGDMA hydrogels, which was mitigated by setpoint correction and fuzzy logic control. SEM and DMA studies showed that the network structure and storage modulus of PEGDMA hydrogels were dependent on the cure time and temporal profile of UV exposure during photopolymerization. This work provides an advance in pulsed and continuous photopolymerization processes for hydrogel engineering based on closed-loop control that enables reproducible fabrication of hydrogels with controlled mechanical properties.

Therapeutic effects of peripherally administrated neural crest stem cells on pain and spinal cord changes after sciatic nerve transection.

BACKGROUND: Severe peripheral nerve injury significantly affects patients' quality of life and induces neuropathic pain. Neural crest stem cells (NCSCs) exhibit several attractive characteristics for cell-based therapies following peripheral nerve injury. Here, we investigate the therapeutic effect of NCSC therapy and associated changes in the spinal cord in a sciatic nerve transection (SNT) model. METHODS: Complex sciatic nerve gap injuries in rats were repaired with cell-free and cell-laden nerve scaffolds for 12 weeks (scaffold and NCSC groups, respectively). Catwalk gait analysis was used to assess the motor function recovery. The mechanical withdrawal threshold and thermal withdrawal latency were used to assess the development of neuropathic pain. Activation of glial cells was examined by immunofluorescence analyses. Spinal levels of extracellular signal-regulated kinase (ERK), NF-κB P65, brain-derived neurotrophic factor (BDNF), growth-associated protein (GAP)-43, calcitonin gene-related peptide (CGRP), and inflammation factors were calculated by western blot analysis. RESULTS: Catwalk gait analysis showed that animals in the NCSC group exhibited a higher stand index and Max intensity At (%) relative to those that received the cell-free scaffold (scaffold group) (p < 0.05). The mechanical and thermal allodynia in the medial-plantar surface of the ipsilateral hind paw were significantly relieved in the NCSC group. Sunitinib (SNT)-induced upregulation of glial fibrillary acidic protein (GFAP) (astrocyte) and ionized calcium-binding adaptor molecule 1 (Iba-1) (microglia) in the ipsilateral L4-5 dorsal and ventral horn relative to the contralateral side. Immunofluorescence analyses revealed decreased astrocyte and microglia activation. Activation of ERK and NF-κB signals and expression of transient receptor potential vanilloid 1 (TRPV1) expression were downregulated. CONCLUSION: NCSC-laden nerve scaffolds mitigated SNT-induced neuropathic pain and improved motor function recovery after sciatic nerve repair. NCSCs also protected the spinal cord from SNT-induced glial activation and central sensitization.

Electrochemical biosensors for pathogen detection.

Recent advances in electrochemical biosensors for pathogen detection are reviewed. Electrochemical biosensors for pathogen detection are broadly reviewed in terms of transduction elements, biorecognition elements, electrochemical techniques, and biosensor performance. Transduction elements are discussed in terms of electrode material and form factor. Biorecognition elements for pathogen detection, including antibodies, aptamers, and imprinted polymers, are discussed in terms of availability, production, and immobilization approach. Emerging areas of electrochemical biosensor design are reviewed, including electrode modification and transducer integration. Measurement formats for pathogen detection are classified in terms of sample preparation and secondary binding steps. Applications of electrochemical biosensors for the detection of pathogens in food and water safety, medical diagnostics, environmental monitoring, and bio-threat applications are highlighted. Future directions and challenges of electrochemical biosensors for pathogen detection are discussed, including wearable and conformal biosensors, detection of plant pathogens, multiplexed detection, reusable biosensors for process monitoring applications, and low-cost, disposable biosensors.

3D bioprinting using hollow multifunctional fiber impedimetric sensors.

3D bioprinting is an emerging biofabrication process for the production of adherent cell-based products, including engineered tissues and foods. While process innovations are rapidly occurring in the area of process monitoring, which can improve fundamental understanding of process-structure-property relations as well as product quality by closed-loop control techniques, in-line sensing of the bioink composition remains a challenge. Here, we report that hollow multifunctional fibers enable in-line impedimetric sensing of bioink composition and exhibit selectivity for real-time classification of cell type, viability, and state of differentiation during bioprinting. Continuous monitoring of the fiber impedance magnitude and phase angle response from 102 to 106 Hz during microextrusion 3D bioprinting enabled compositional and quality analysis of alginate bioinks that contained fibroblasts, neurons, or mouse embryonic stem cells (mESCs). Fiber impedimetric responses associated with the bioinks that contained differentiated mESCs were consistent with differentiation marker expression characterized by immunocytochemistry. 3D bioprinting through hollow multifunctional fiber impedimetric sensors enabled classification of stem cells as stable or randomly differentiated populations. This work reports an advance in monitoring 3D bioprinting processes in terms of in-line sensor-based bioink compositional analysis using fiber technology and provides a non-invasive sensing platform for achieving future quality-controlled bioprinted tissues and injectable stem-cell therapies.

3D Printed Multiplexed Competitive Migration Assays with Spatially Programmable Release Sources.

Here, a 3D printed multiplexed competitive migration assay is reported for characterizing a chemotactic response in the presence of multiple spatially distributed chemoattractants. The utility of the assay is demonstrated by examining the chemotactic response of human glioblastoma cells to spatially opposing chemotactic gradients of epidermal growth factor (EGF) and bradykinin (BK). Competitive migration assays involving spatially opposing gradients of EGF and BK that are optimized in the absence of the second chemoattractant show that 46% more glioblastoma cells migrate toward EGF sources. The migration velocities of human glioblastoma cells toward EGF and BK sources are reduced by 7.6 ± 2.2% and 11.6 ± 6.3% relative to those found in the absence of the spatially opposing chemoattractant. This work provides new insight to the chemotactic response associated with glioblastoma-vasculature interactions and a versatile, user-friendly platform for characterizing the chemotactic response of cells in the presence of multiple spatially distributed chemoattractants.

Corrigendum to "3D printed bionic nanodevices" [Nano Today 11 (2016) 330-350].

[This corrects the article PMC5016035.].

Low-cost sensor-integrated 3D-printed personalized prosthetic hands for children with amniotic band syndrome: A case study in sensing pressure distribution on an anatomical human-machine interface (AHMI) using 3D-printed conformal electrode arrays.

Interfacing anatomically conformal electronic components, such as sensors, with biology is central to the creation of next-generation wearable systems for health care and human augmentation applications. Thus, there is a need to establish computer-aided design and manufacturing methods for producing personalized anatomically conformal systems, such as wearable devices and human-machine interfaces (HMIs). Here, we show that a three-dimensional (3D) scanning and 3D printing process enabled the design and fabrication of a sensor-integrated anatomical human-machine interface (AHMI) in the form of personalized prosthetic hands that contain anatomically conformal electrode arrays for children affected by amniotic band syndrome, a common birth defect. A methodology for identifying optimal scanning parameters was identified based on local and global metrics of registered point cloud data quality. This method identified an optimal rotational angle step size between adjacent 3D scans. The sensitivity of the optimization process to variations in organic shape (i.e., geometry) was examined by testing other anatomical structures, including a foot, an ear, and a porcine kidney. We found that personalization of the prosthetic interface increased the tissue-prosthesis contact area by 408% relative to the non-personalized devices. Conformal 3D printing of carbon nanotube-based polymer inks across the personalized AHMI facilitated the integration of electronic components, specifically, conformal sensor arrays for measuring the pressure distribution across the AHMI (i.e., the tissue-prosthesis interface). We found that the pressure across the AHMI exhibited a non-uniform distribution and became redistributed upon activation of the prosthetic hand's grasping action. Overall, this work shows that the integration of 3D scanning and 3D printing processes offers the ability to design and fabricate wearable systems that contain sensor-integrated AHMIs.

Additive Manufacturing of Mechanically Isotropic Thin Films and Membranes via Microextrusion 3D Printing of Polymer Solutions.

Polymer extrusion additive manufacturing processes, such as fused filament fabrication (FFF), are now being used to explore the fabrication of thin films and membranes. However, the physics of molten polymer extrusion constrains achievable thin film properties (e.g., mechanical isotropy), material selection, and spatial control of film composition. Herein, we present an approach for fabrication of functional polymer thin films and membranes based on the microextrusion printing of polymer solutions, which we refer to as "solvent-cast printing" (SCP). Constructs fabricated via SCP exhibited a 43% reduction in anisotropy of tensile strength relative to those fabricated using FFF. The constructs fabricated via SCP exhibited a lesser extent of visible layering defects relative to those fabricated by FFF. Further, the swelling dynamics of the films varied depending on the membrane fabrication technique (i.e., SCP vs manual drop casting). The opportunity for expanding material selection relative to FFF processes was demonstrated by printing poly(benzimidazole), a high-performance thermoplastic with high glass-transition temperatures ( Tg ∼ 400 °C). Results from this work indicate that our new approach could facilitate the manufacture of mechanically isotropic thin films and membranes using currently unprintable high-performance thermoplastics.

Process- and bio-inspired hydrogels for 3D bioprinting of soft free-standing neural and glial tissues.

A bio-inspired hydrogel for 3D bioprinting of soft free-standing neural tissues is presented. The novel filler-free bioinks were designed by combining natural polymers for extracellular matrix biomimicry with synthetic polymers to endow desirable rheological properties for 3D bioprinting. Crosslinking of thiolated Pluronic F-127 with dopamine-conjugated (DC) gelatin and DC hyaluronic acid through a thiol-catechol reaction resulted in thermally gelling bioinks with Herschel-Bulkley fluid rheological behavior. Microextrusion 3D bioprinting was used to fabricate free-standing cell-laden tissue constructs. The bioinks exhibited flattened parabolic velocity profiles with tunable low shear regions. Two pathways were investigated for curing the bioink: chelation and photocuring. The storage modulus of the cured bioinks ranged from 6.7 to 11.7 kPa. The iron (III) chelation chemistry produced crosslinked neural tissues of relatively lower storage modulus than the photocuring approach. In vitro cell viability studies using the 3D bioprinted neural tissues showed that the cured bioink was biocompatible based on minimal cytotoxic response observed over seven days in culture relative to control studies using alginate hydrogels. Rodent Schwann cell-, rodent neuronal cell-, and human glioma cell-laden tissue constructs were printed and cultured over seven days and exhibited comparable viability relative to alginate bioink controls. The ability to fabricate soft, free-standing 3D neural tissues with low modulus has implications in the biofabrication of microphysiological neural systems for disease modeling as well as neural tissues and innervated tissues for regenerative medicine.

Three-dimensional (3D) printed scaffold and material selection for bone repair.

Critical-sized bone defect repair remains a substantial challenge in clinical settings and requires bone grafts or bone substitute materials. However, existing biomaterials often do not meet the clinical requirements of structural support, osteoinductive property, and controllable biodegradability. To treat large-scale bone defects, the development of three-dimensional (3D) porous scaffolds has received considerable focus within bone engineering. A variety of biomaterials and manufacturing methods, including 3D printing, have emerged to fabricate patient-specific bioactive scaffolds that possess controlled micro-architectures for bridging bone defects in complex configurations. During the last decade, with the development of the 3D printing industry, a large number of tissue-engineered scaffolds have been created for preclinical and clinical applications using novel materials and innovative technologies. Thus, this review provides a brief overview of current progress in existing biomaterials and tissue engineering scaffolds prepared by 3D printing technologies, with an emphasis on the material selection, scaffold design optimization, and their preclinical and clinical applications in the repair of critical-sized bone defects. Furthermore, it will elaborate on the current limitations and potential future prospects of 3D printing technology. STATEMENT OF SIGNIFICANCE: 3D printing has emerged as a critical fabrication process for bone engineering due to its ability to control bulk geometry and internal structure of tissue scaffolds. The advancement of bioprinting methods and compatible ink materials for bone engineering have been a major focus to develop optimal 3D scaffolds for bone defect repair. Achieving a successful balance of cellular function, cellular viability, and mechanical integrity under load-bearing conditions is critical. Hybridization of natural and synthetic polymer-based materials is a promising approach to create novel tissue engineered scaffolds that combines the advantages of both materials and meets various requirements, including biological activity, mechanical strength, easy fabrication and controllable degradation. 3D printing is linked to the future of bone grafts to create on-demand patient-specific scaffolds.

Programming of Multicomponent Temporal Release Profiles in 3D Printed Polypills via Core-Shell, Multilayer, and Gradient Concentration Profiles.

Additive manufacturing (AM) appears poised to provide novel pharmaceutical technology and controlled release systems, yet understanding the effects of processing and post-processing operations on pill design, quality, and performance remains a significant barrier. This paper reports a study of the relationship between programmed concentration profile and resultant temporal release profile using a 3D printed polypill system consisting of a Food and Drug Administration (FDA) approved excipient (Pluronic F-127) and therapeutically relevant dosages of three commonly used oral agents for treatment of type 2 diabetes (300-500 mg per pill). A dual-extrusion hydrogel microextrusion process enables the programming of three unique concentration profiles, including core-shell, multilayer, and gradient structures. Experimental and computational studies of diffusive mass transfer processes reveal that programmed concentration profiles are dynamic throughout both pill 3D printing and solidification. Spectrophotometric assays show that the temporal release profiles could be selectively programmed to exhibit delayed, pulsed, or constant profiles over a 5 h release period by utilizing the core-shell, multilayer, and gradient distributions, respectively. Ultimately, this work provides new insights into the mass transfer processes that affect design, quality, and performance of spatially graded controlled release systems, as well as demonstrating the potential to create disease-specific polypill technology with programmable temporal release profiles.

Additive manufacturing of three-dimensional (3D) microfluidic-based microelectromechanical systems (MEMS) for acoustofluidic applications.

Three-dimensional (3D) printing now enables the fabrication of 3D structural electronics and microfluidics. Further, conventional subtractive manufacturing processes for microelectromechanical systems (MEMS) relatively limit device structure to two dimensions and require post-processing steps for interface with microfluidics. Thus, the objective of this work is to create an additive manufacturing approach for fabrication of 3D microfluidic-based MEMS devices that enables 3D configurations of electromechanical systems and simultaneous integration of microfluidics. Here, we demonstrate the ability to fabricate microfluidic-based acoustofluidic devices that contain orthogonal out-of-plane piezoelectric sensors and actuators using additive manufacturing. The devices were fabricated using a microextrusion 3D printing system that contained integrated pick-and-place functionality. Additively assembled materials and components included 3D printed epoxy, polydimethylsiloxane (PDMS), silver nanoparticles, and eutectic gallium-indium as well as robotically embedded piezoelectric chips (lead zirconate titanate (PZT)). Electrical impedance spectroscopy and finite element modeling studies showed the embedded PZT chips exhibited multiple resonant modes of varying mode shape over the 0-20 MHz frequency range. Flow visualization studies using neutrally buoyant particles (diameter = 0.8-70 µm) confirmed the 3D printed devices generated bulk acoustic waves (BAWs) capable of size-selective manipulation, trapping, and separation of suspended particles in droplets and microchannels. Flow visualization studies in a continuous flow format showed suspended particles could be moved toward or away from the walls of microfluidic channels based on selective actuation of in-plane or out-of-plane PZT chips. This work suggests additive manufacturing potentially provides new opportunities for the design and fabrication of acoustofluidic and microfluidic devices.