Pedro Huebner

Regenerative medicine approaches to restore the mandibular condyle of the temporomandibular joint (TMJ) may fill an unmet patient need. In this study, a method to implant an acellular regenerative TMJ prosthesis was developed for orthotopic implantation in a pilot goat study. The scaffold incorporated a porous, polycaprolactone-hydroxyapatite (PCL-HAp, 20wt% HAp) 3D printed condyle with a cartilage-matrix-containing hydrogel. A series of material characterizations was used to determine the structure, fluid transport, and mechanical properties of 3D printed PCL-HAp. To promote marrow uptake for cell seeding, a scaffold pore size of 152 ± 68 μm resulted in a whole blood transport initial velocity of 3.7 ± 1.2 mm·s ⁻¹ transported to the full 1 cm height. The Young's modulus of PCL was increased by 67% with the addition of HAp, resulting in a stiffness of 269 ± 20 MPa for etched PCL-HAp. In addition, the bending modulus increased by 2.06-fold with the addition of HAp to 470 MPa for PCL-HAp. The prosthesis design with an integrated hydrogel was compared with unoperated contralateral control and no-hydrogel group in a goat model for 6 months. A guide was used to make the condylectomy cut, and the TMJ disc was preserved. MicroCT assessment of bone suggested variable tissue responses with some regions of bone growth and loss, although more loss may have been exhibited by the hydrogel group than the no-hydrogel group. A benchtop load transmission test suggested that the prosthesis was not shielding load to the underlying bone. Although variable, signs of neocartilage formation were exhibited by Alcian blue and collagen II staining on the anterior, functional surface of the condyle. Overall, this study demonstrated signs of functional TMJ restoration with an acellular prosthesis. There were apparent limitations to continuous, reproducible bone formation, and stratified zonal cartilage regeneration. Future work may refine the prosthesis design for a regenerative TMJ prosthesis amenable to clinical translation.

Experimental adhesives with functional nitrogen-doped titanium dioxide nanoparticles (N_TiO₂) have been shown to display improved properties. However, these materials have not been characterized regarding their degree of conversion (DC), biaxial flexure strength (BFS), surface roughness (SR), elastic modulus (EM), and long-term antibacterial functionalities. Experimental adhesives were synthesized by dispersing N_TiO₂ (10%, 20%, or 30%, v/v%) into OptiBond Solo Plus (OPTB, Kerr Corp., USA). Unpolymerized adhesives (volume = 50 μL/drop, n = 3/group) were individually placed onto a heated (37 °C) attenuated total reflectance (ATR) monolithic diamond crystal (Golden Gate, Specac). The spectra of composites were obtained with a Fourier-transform infrared (FTIR) spectrometer (Nicolet IS50; 500–4500 cm⁻¹; resolution = 4 cm⁻¹, 10 internal scans/spectrum) before and after polymerization. Disk-shaped specimens (diameter = 6.0 mm, thickness = 0.5 mm) for BFS (n = 12/group), SR and EM (n = 3/group), and for antibacterial testing (n = 18/group/time-point) were fabricated and photopolymerized (1 min each; 385–515 nm, 1000 mW/cm²; VALO). DC values (%) were calculated from pre- and post-polymerization spectra using the two-frequency method and tangent-baseline technique. BFS was assessed using a universal testing machine (Instron 68TM-5, crosshead speed = 1.27 mm/min, 25 °C). SR and EM were investigated using an atomic force microscope (Multimode 8) with aluminum-coated silicon probes (8 nm pyramidal tip, spring constant 40 N/m, Bruker). Antibacterial testing was performed by growing Streptococcus mutans biofilms (UA159-ldh, 37 °C, microaerophilic) on the surfaces of specimens for 24 h and then measuring the relative luminescence units (RLU) with a Biotek Synergy HT multi-well plate reader. Results demonstrate that experimental materials containing 10%, 20%, and 30% of N_TiO₂ displayed higher levels of DC, had better mechanical properties, and were able to exert strong and durable antibacterial properties without visible light irradiation and after extended periods of simulated shelf-life and aging in PBS. The reported experimental materials are expected to increase the service lives of polymer-based bonded restorations by decreasing the incidence of secondary caries.

The COVID-19 global pandemic caused by the widespread transmission of the novel coronavirus (SARS-CoV-2) has become one of modern history’s most challenging issues from a healthcare perspective. At its dawn, still without a vaccine, contagion containment strategies remained most effective in preventing the disease’s spread. Patient isolation has been primarily driven by the results of polymerase chain reaction (PCR) testing, but its initial reach was challenged by low availability and high cost, especially in developing countries. As a means of taking advantage of a preexisting infrastructure for respiratory disease diagnosis, researchers have proposed COVID-19 patient screening based on the results of Chest Computerized Tomography (CT) and Chest Radiographs (X-ray). When paired with artificial-intelligence- and deep-learning-based approaches for analysis, early studies have achieved a comparatively high accuracy in diagnosing the disease. Considering the opportunity to further explore these methods, we implement six different Deep Convolutional Neural Networks (Deep CNN) models—VGG16, MobileNetV2, InceptionResNetV2, ResNet50, ResNet101, and VGG19—and use a mixed dataset of CT and X-ray images to classify COVID-19 patients. Preliminary results showed that a modified MobileNetV2 model performs best with an accuracy of 95 ± 1.12% (AUC = 0.816). Notably, a high performance was also observed for the VGG16 model, outperforming several previously proposed models with an accuracy of 98.5 ± 1.19% on the X-ray dataset. Our findings are supported by recent works in the academic literature, which also uphold the higher performance of MobileNetV2 when X-ray, CT, and their mixed datasets are considered. Lastly, we further explain the process of feature extraction using Local Interpretable Model-Agnostic Explanations (LIME), which contributes to a better understanding of what features in CT/X-ray images characterize the onset of COVID-19.

Speech perception in an adverse background/noisy environment is a complex and challenging human process, which is made even more complicated in foreign-accented language for bilingual and monolingual individuals. Listeners who have difficulties in hearing are affected most by such a situation. Despite considerable efforts, the increase in speech intelligibility in noise remains elusive. Considering this opportunity, this study investigates Bengali–English bilinguals and native American English monolinguals’ behavioral patterns on foreign-accented English language considering bubble noise, gaussian or white noise, and quiet sound level. Twelve regular hearing participants (Six Bengali–English bilinguals and Six Native American English monolinguals) joined in this study. Statistical computation shows that speech with different noise has a significant effect (p = 0.009) on listening for both bilingual and monolingual under different sound levels (e.g., 55 dB, 65 dB, and 75 dB). Here, six different machine learning approaches (Logistic Regression (LR), Linear Discriminant Analysis (LDA), K-nearest neighbors (KNN), Naïve Bayes (NB), Classification and regression trees (CART), and Support vector machine (SVM)) are tested and evaluated to differentiate between bilingual and monolingual individuals from their behavioral patterns in both noisy and quiet environments. Results show that most optimal performances were observed using LDA by successfully differentiating between bilingual and monolingual 60% of the time. A deep neural network-based model is proposed to improve this measure further and achieved an accuracy of nearly 100% in successfully differentiating between bilingual and monolingual individuals.

Chest radiographs (X-rays) combined with Deep Convolutional Neural Network (CNN) methods have been demonstrated to detect and diagnose the onset of COVID-19, the disease caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). However, questions remain regarding the accuracy of those methods as they are often challenged by limited datasets, performance legitimacy on imbalanced data, and have their results typically reported without proper confidence intervals. Considering the opportunity to address these issues, in this study, we propose and test six modified deep learning models, including VGG16, InceptionResNetV2, ResNet50, MobileNetV2, ResNet101, and VGG19 to detect SARS-CoV-2 infection from chest X-ray images. Results are evaluated in terms of accuracy, precision, recall, and f- score using a small and balanced dataset (Study One), and a larger and imbalanced dataset (Study Two). With 95% confidence interval, VGG16 and MobileNetV2 show that, on both datasets, the model could identify patients with COVID-19 symptoms with an accuracy of up to 100%. We also present a pilot test of VGG16 models on a multi-class dataset, showing promising results by achieving 91% accuracy in detecting COVID-19, normal, and Pneumonia patients. Furthermore, we demonstrated that poorly performing models in Study One (ResNet50 and ResNet101) had their accuracy rise from 70% to 93% once trained with the comparatively larger dataset of Study Two. Still, models like InceptionResNetV2 and VGG19's demonstrated an accuracy of 97% on both datasets, which posits the effectiveness of our proposed methods, ultimately presenting a reasonable and accessible alternative to identify patients with COVID-19.

The limitations and high false-negative rates (30%) of COVID-19 test kits have been a prominent challenge during the 2020 coronavirus pandemic. Manufacturing those kits and performing the tests require extensive resources and time. Recent studies show that radiological images like chest X-rays can offer a more efficient solution and faster initial screening of COVID-19 patients. In this study, we develop a COVID-19 diagnosis model using Multilayer Perceptron and Convolutional Neural Network (MLP-CNN) for mixed-data (numerical/categorical and image data). The model predicts and differentiates between COVID-19 and non-COVID-19 patients, such that early diagnosis of the virus can be initiated, leading to timely isolation and treatments to stop further spread of the disease. We also explore the benefits of using numerical/categorical data in association with chest X-ray images for screening COVID-19 patients considering both balanced and imbalanced datasets. Three different optimization algorithms are used and tested:adaptive learning rate optimization algorithm (Adam), stochastic gradient descent (Sgd), and root mean square propagation (Rmsprop). Preliminary computational results show that, on a balanced dataset, a model trained with Adam can distinguish between COVID-19 and non-COVID-19 patients with a higher accuracy of 96.3%. On the imbalanced dataset, the model trained with Rmsprop outperformed all other models by achieving an accuracy of 95.38%. Additionally, our proposed model outperformed selected existing deep learning models (considering only chest X-ray or CT scan images) by producing an overall average accuracy of 94.6% ± 3.42%.

Osteoarthritis is a degenerative joint disease that limits mobility of the affected joint due to the degradation of articular cartilage and subchondral bone. The limited regenerative capacity of cartilage presents significant challenges when attempting to repair or reverse the effects of cartilage degradation. Tissue engineered medical products are a promising alternative to treat osteochondral degeneration due to their potential to integrate into the patient's existing tissue. The goal of this study was to create a scaffold that would induce site‐specific osteogenic and chondrogenic differentiation of human adipose‐derived stem cells (hASC) to generate a full osteochondral implant. Scaffolds were fabricated using 3D‐bioplotting of biodegradable polycraprolactone (PCL) with either β‐tricalcium phosphate (TCP) or decellularized bovine cartilage extracellular matrix (dECM) to drive site‐specific hASC osteogenesis and chondrogenesis, respectively. PCL‐dECM scaffolds demonstrated elevated matrix deposition and organization in scaffolds seeded with hASC as well as a reduction in collagen I gene expression. 3D‐bioplotted PCL scaffolds with 20% TCP demonstrated elevated calcium deposition, endogenous alkaline phosphatase activity, and osteopontin gene expression. Osteochondral scaffolds comprised of hASC‐seeded 3D‐bioplotted PCL‐TCP, electrospun PCL, and 3D‐bioplotted PCL‐dECM phases were evaluated and demonstrated site‐specific osteochondral tissue characteristics. This technique holds great promise as cartilage morbidity is minimized since autologous cartilage harvest is not required, tissue rejection is minimized via use of an abundant and accessible source of autologous stem cells, and biofabrication techniques allow for a precise, customizable methodology to rapidly produce the scaffold.

Purpose: Musculoskeletal soft tissues possess highly aligned extracellular collagenous networks that provide structure and strength. Such an organization dictates tissue-specific mechanical properties but can be difficult to replicate by engineered biological substitutes. Nanofibrous electrospun scaffolds have demonstrated the ability to control cell-secreted collagen alignment, but concerns exist regarding their scalability for larger and anatomically relevant applications. Additive manufacturing processes, such as melt extrusion-based 3D-Bioplotting, allow fabrication of structurally relevant scaffolds featuring highly controllable porous microarchitectures. Materials and Methods: In this study, we investigate the effects of 3D-bioplotted scaffold design on the compressive elastic modulus of neotissue formed in vivo in a subcutaneous rat model and its correlation with the alignment of ECM collagen fibers. Polycaprolactone scaffolds featuring either 100 or 400 µm interstrand spacing were implanted for 4 or 12 weeks, harvested, cryosectioned, and characterized using atomic-force-microscopy-based force mapping. Results: The compressive elastic modulus of the neotissue formed within the 100 µm design was significantly higher at 4 weeks (p < 0.05), but no differences were observed at 12 weeks. In general, the tissue stiffness was within the same order of magnitude and range of values measured in native musculoskeletal soft tissues including the porcine meniscus and anterior cruciate ligament. Finally, a significant positive correlation was noted between tissue stiffness and the degree of ECM collagen fiber alignment (p < 0.05) resulting from contact guidance provided by scaffold strands. Conclusion: These findings demonstrate the significant effects of 3D-bioplotted scaffold microarchitectures in the organization and sub-tissue-level mechanical properties of ECM in vivo.

Purpose: Tissue engineering and regenerative medicine approaches have the potential to overcome the challenges associated with current treatment strategies for meniscus injuries. 3D-Bioplotted scaffolds are promising, but have not demonstrated the ability to guide the formation of aligned collagenous matrix in vivo, which is critical for generating functional meniscus tissue. In this study, we evaluate the ability of 3D-Bioplotted scaffold designs with varying interstrand spacing to induce the deposition of aligned matrix in vivo. Materials and Methods: 3D-Bioplotted polycaprolactone scaffolds with 100, 200, or 400 μm interstrand spacing were implanted subcutaneously in a rat model for 4, 8, or 12 weeks. Scaffolds were harvested, paraffin-embedded, sectioned, and stained to visualize cell nuclei and collagen. Quantitative image analysis was used to evaluate cell density, matrix fill, and collagen fiber alignment within the scaffolds. Results: By 4 weeks, cells had infiltrated the innermost scaffold regions. Similarly, collagenous matrix filled interstrand regions nearly completely by 4 weeks. By 12 weeks, aligned collagen was present in all scaffolds. Generally, alignment along the scaffold strands increased over time for all three interstrand spacing groups. Distribution of collagen fiber alignment angles narrowed as interstrand spacing decreased. Conclusions: 3D-Bioplotted scaffolds allow for complete cell infiltration and collagenous matrix production throughout the scaffold. The ability to use interstrand spacing as a means of controlling the formation of aligned collagen in vivo was demonstrated, which helps establish a design space for scaffold-based meniscus tissue engineering.

Electrospun scaffolds provide a dense framework of nanofibers with pore sizes and fiber diameters that closely resemble the architecture of native extracellular matrix. However, it generates limited three-dimensional structures of relevant physiological thicknesses. 3D printing allows digitally controlled fabrication of three-dimensional single/multimaterial constructs with precisely ordered fiber and pore architecture in a single build. However, this approach generally lacks the ability to achieve submicron resolution features to mimic native tissue. The goal of this study was to fabricate and evaluate 3D printed, electrospun, and combination of 3D printed/electrospun scaffolds to mimic the native architecture of heterogeneous tissue. We assessed their ability to support viability and proliferation of human adipose derived stem cells (hASC). Cells had increased proliferation and high viability over 21 days on all scaffolds. We further tested implantation of stacked-electrospun scaffold versus combined electrospun/3D scaffold on a cadaveric pig knee model and found that stacked-electrospun scaffold easily delaminated during implantation while the combined scaffold was easier to implant. Our approach combining these two commonly used scaffold fabrication technologies allows for the creation of a scaffold with more close resemblance to heterogeneous tissue architecture, holding great potential for tissue engineering and regenerative medicine applications of osteochondral tissue and other heterogeneous tissues.