Search

Recent advances in deep learning have provided an opportunity to improve and automate dysarthria intelligibility assessment, offering a cost-effective, accessible, and less subjective way to assess dysarthric speakers. However, reviewing previous literature in the area determines that the generalization of results on new dysarthric patients was not measured properly or incomplete among the previous studies that yielded very high accuracies due to the gaps in the adopted evaluation methodologies. This is of particular importance as any practical and clinical application of intelligibility assessment approaches must reliably generalize on new patients; otherwise, the clinicians cannot accept the assessment results provided by the system deploying the approach. In this paper, after these gaps are explained, we report on our extensive investigation to propose a deep learning–based dysarthric intelligibility assessment optimal setup. Then, we explain different evaluation strategies that were applied to thoroughly verify how the optimal setup performs with new speakers and across different classes of speech intelligibility. Finally, a comparative study was conducted, benchmarking the performance of our proposed optimal setup against the state of the art by adopting similar strategies previous studies employed. Results indicate an average of 78.2% classification accuracy for unforeseen low intelligibility speakers, 40.6% for moderate intelligibility speakers, and 40.4% for high intelligibility speakers. Furthermore, we noticed a high variance of classification accuracies among individual speakers. Finally, our proposed optimal setup delivered an average of 97.19% classification accuracy when adopting a similar evaluation strategy used by the previous studies.

Our knowledge of the disease burden and symptoms with age in COVID-19 patients is limited. Therefore, it is of interest to document the clinical aspect of this association with respect to the disease. We used the data of 3363 patients enrolled with an urgent care clinic in Volusia county, Florida for this study. Data shows difference in age among COVID-19 antibody (Ab) - positive patients (48.3 years, 95% CI = 46.9, 49.7 years) and Ab-negative patients (46.1 years, 95% CI = 45.4, 46.8 years). However, disease burden by age is not significant on average. Nonetheless, COVID-19 positive patients between 40-69-years of age experienced the highest burden of disease and highest average number of symptoms. Thus, COVID-19 disease burden and number of symptoms experienced were highest among the 40-69-year-old patients. Those above the populations mean age of 46.4 years old were more likely to test positive for COVID-19.

Laparoscopic surgery has a notably high learning curve, hindering typical approaches to training. Due to unique challenges that are not present in open surgery (the hinge effect, small field of view (FoV), lack of depth perception, and small workspace), a surgical resident may be delayed in participating in laparoscopic surgery until later in residency. Having a narrow window to complete highly specialized training can lead to graduates feeling under-prepared for solo practice. Additionally, delayed introduction may expose trainees to fewer than 200 laparoscopic cases. Therefore, there is a need for surgical residents to increase both their caseload and training window without compromising patient safety. This project aims to develop and test a proof-of-concept prototype that uses granular jamming technology to controllably vary the force required to move a laparoscopic tool. By increasing tool resistance, the device helps prevents accidental injury to important nearby anatomical structures such as urinary tract, vasculature, and/or bowel. Increasing the safety of laparoscopic surgery would allow residents to begin their training earlier, gaining exposure and confidence. A device to adjust tool resistance has benefits to the experienced surgeon as well – surgeries require continuous tool adjustment and tension, resulting in fatigue. Increasing tool resistance can assist surgeons in situations requiring continuous tension and can also provide safety against sudden movements. This investigational device was prototyped using SolidWorks CAD software, then 3D printed and assessed with a laparoscopic box trainer.

This paper addresses the optimal allocation of battery energy storage systems (BESS) in radial distribution systems for bus voltage regulation and energy cost reduction. A hierarchical planning model is proposed to obtain the optimal configuration (number, locations, and sizes) of BESS. The three interacting levels in the proposed model are: (1) determination of the optimal BESS number; (2) use of a heuristic strategy based on voltage sensitivity analysis to select the most effective locations for a given BESS number in the network; and (3) sizing the BESS capacity to realize optimal operation objectives. The final BESS configuration is selected based on minimization of total life cycle cost, which takes into account the investment cost, residual value of BESS, and system operation cost. The voltage regulation objectives are modeled as constraints. The proposed method reduces computational burden significantly and provides a comprehensive sensitivity analysis and locate the most effective buses to mitigate voltage violation. The effectiveness of this optimal allocation method was demonstrated for a modified IEEE 15-bus and IEEE 69-bus distribution network.

The current structural engineering practical standards are unable to offer an universal structural design standard for long-spanning lightweight stadium roofing structures. As such, the design procedure of a particular stadium roof is not replicable to another. This research aims to present a novel design procedure for lightweight stadium roofing structures considering the Lakhwiya stadium the Optus Stadium and the CommBank Stadium as experimental cases. Using the finite element analysis (FEA) software Strand7, the cases will be modelled and analysed. Varying load cases and combinations such as ultimate strength (ULS) and serviceability limit states (SLS) based on the Australian Standard AS1170.0:2002 will be calculated and subsequently applied. Linear static analysis will then be undertaken where critical members will be identified within the model. Based on this, preliminary member sizing and design feasibility checks will be conducted in order to ensure structural stability and compliance to the Australian Steel Structure code AS4100:2020. A linear buckling analysis is also conducted based on the selected sizes from the initial stage to determine critical loads. Advanced analysis including non-linear buckling computation is comprehensively managed. Some of the crucial parameters such as maximum displacement, maximum/minimum principal stresses, critical buckling loads, as well as load factors are examined. The main novelty of this study is to determine a clear road map to design stadium roofing systems subjected to a combination of different types of the loading.

Automated dysarthria intelligibility assessment offers the opportunity to develop reliable, low-cost, and scalable tools, which help to solve current shortcomings of manual and subjective intelligibility assessments. This paper reviews the literature regarding automated intelligibility assessment, identifying the highest performing published models and concluding on promising avenues for further research. Our review shows that most of the existing work were able to achieve very high accuracies. However, we have found that most of these studies validated their models using speech samples of the same speakers used in training, making their results less generalizable. Furthermore, there is a lack of study on how well these models perform on speakers from different datasets or different microphone setups. This lack of generalizability has implications to the real-life application of these models. Future research directions could include the use of more robust methods of validation such as using unseen speakers, as well as incorporating speakers from different datasets. This would provide confidence that the models are generalized and therefore allow them to be used in real-world clinical practice.

We present a Physics-Informed Graph Neural Network (pigNN) methodology for rapid and automated compact model development. It brings together the inherent strengths of data-driven machine learning, high-fidelity physics in TCAD simulations, and knowledge contained in existing compact models. In this work, we focus on developing a neural network (NN) based compact model for a non-ideal PN diode that represents one nonlinear edge in a pigNN graph. This model accurately captures the smooth transition between the exponential and quasi-linear response regions. By learning voltage dependent non-ideality factor using NN and employing an inverse response function in the NN loss function, the model also accurately captures the voltage dependent recombination effect. This NN compact model serves as basis model for a PN diode that can be a single device or represent an isolated diode in a complex device determined by topological data analysis (TDA) methods. The pigNN methodology is also applicable to derive reduced order models in other engineering areas.