70040 research outputs found
Sort by
Hairpin-locker mediated CRISPR/Cas tandem system for ultrasensitive detection of DNA without pre-amplification
Full text linkAchieving ultra-sensitive detection of DNA is of paramount importance in the field of molecular analytics. Conventional amplification technologies such as polymerase chain reaction (PCR) currently play a leading role in ultrasensitive DNA detection. However, amplicon contamination common in these techniques may lead to false positives. To date, CRISPR-associated nucleases (type V & VI) with their programmable cleavage have been utilised for sensitive detection of unamplified nucleic acids in complex real samples. Nevertheless, without additional amplification strategies, the pM range sensitivity of such CRISPR/Cas sensors is not sufficient for clinical applications. Here, we established a hairpin-locker (H-locker) mediated Cas12-Cas13 tandem biosensing system (Cas12-13 tandem-sensor) for ultrasensitive detection of DNA targets. Without the need for any additional amplification reaction or device, this system is capable of detecting DNA at a notable 1 aM level (<1 copy/μL) sensitivity. In addition, the system was able to distinguish cancer mutations in colorectal cancer (CRC) mice. This is a significant advance for CRISPR/Cas biosensing technology offering simple, highly sensitive, and user-friendly diagnostics for next-generation nucleic acid detection
Deep learning-based skin cancer diagnosis
No full textSkin cancer is the most common type of cancer worldwide. Traditional diagnostic methods based on dermatologists examining samples under a microscope are time intensive and subjective, with notable variability between and within individual dermatologist evaluations. The advent of deep learning and advances in computational hardware have led to the development of computer-aided diagnosis (CAD) systems aimed at helping pathologists. However, challenges remain, including the reliance on models designed for natural image datasets that do not address domain-specific differences and the need for accurate and detailed medical report generation from input images.
This thesis presents three research contributions that address these challenges. The first study focuses on diagnosing basal cell carcinoma from whole slide images. By utilizing neural architecture search (NAS), an optimal network was developed specifically for skin cancer detection, outperforming traditional methods. A supernet, SC-net, was introduced to ensure fair training and reduce evaluation bias, leveraging evolutionary search to identify the best architecture. The proposed method demonstrated robust generalizability and effectiveness across both skin cancer datasets and other medical datasets.
The second study explores melanoma detection from clinical dermoscopic images. A novel framework, SCD-NAS, integrates NAS with large language models (LLM) to identify architectures optimized for skin cancer analysis. SCD-NAS offers a more adaptable solution compared to conventional NAS methods, addressing the complexities of skin cancer diagnosis. The framework exhibited superior performance on datasets such as ISIC 2020, MedMNISTv2, CIFAR-10 and CIFAR-100, underscoring its generalizability.
The third study focuses on generating medical reports from medical images. The proposed Universal Medical Report Generation (UniMRG) framework employs vision-language (VL) foundation models to enhance the understanding of semantic relationships in medical image-text pairs. Innovations such as multi-modal data augmentation, visual-textual alignment modules and a Proxy-Guided Iterative search strategy improve the adaptability and performance of the framework. UniMRG proves to be a versatile solution, applicable not only to dermatopathology images but also to other medical images, setting a new standard for automated medical report generation
The impact of local corruption on firms' narrative R&D disclosures
Full text linkThis study examines the impact of local corruption on firms' narrative research and development (R&D) disclosures in the United States. We find that firms in more corrupt areas include fewer R&D sentences in their 10-K filings, and these sentences contain less numerical and forward-looking information. Our results hold across various measures of local corruption and R&D disclosures and remain robust after controlling for firms' R&D activities, implementing fixed effects, using difference-in-differences tests, and applying instrumental variable analysis. Additionally, the effects are more pronounced for firms with concentrated operations in their headquarters states and for firms whose R&D disclosures closely relate to future earnings. However, they are less pronounced for firms with CEOs politically aligned with the state's incumbent party and when the benefit of resolving market dispersion from firms' R&D disclosure is high. Overall, our findings indicate that local corruption adversely affects firms' narrative R&D disclosures
‘Ōlē bābālē!!!(Figure presented)’ – Exploring emotions and digital language practices in family language policy among Bengali transnational families: A multimodal approach
Full text linkAims and objectives: This study explores the role of emotions in heritage language (HL) practice within the context of networked family language policy (nFLP) among multilingual, transnational families through digital communication using multimodal (inter)action analysis (MIA). Methodology: The study employs a case study methodological approach involving Sigrid Norris’s MIA tools to understand how emotions are expressed and shared within multimodal digital interactions and how these interactions shape HL practice in transnational families. Data and analysis: The following data were collected from three Bengali immigrant families in Australia employing (1) participant-led video recordings of online conversations between the child and the grandparent, (2) screenshots of online chats, and (3) semi-structured interview data to support the video and screenshot data. Findings/conclusions: The results show that emotional factors, such as emotional contagion and digitally mediated affective communication, drive modally rich HL practice. Family bonding with transnational grandparents, shaped by these emotional factors, was found to be the primary goal of digital HL communication practice, while HL maintenance was secondary. Originality: This research provides a nuanced and fine-grained analysis of emotional factors in HL practice with transnational grandparents through digital media using a unique toolset (MIA). Significance/implications: This research promotes the significant role of emotional factors in HL practice when it comes to using digital communication among immigrant transnational families
Inhibition of the NLRP3 inflammasome using MCC950 reduces vincristine-induced adverse effects in an acute lymphoblastic leukemia patient-derived xenograft model
Full text linkVincristine is one of the most important chemotherapeutic drugs used to treat acute lymphoblastic leukemia (ALL). Unfortunately, vincristine often causes severe adverse effects, including sensory–motor neuropathies, weight loss, and overall decreased well-being, that are difficult to control and that decrease the quality of life and survival of patients. Recent studies demonstrate that sensory–motor adverse effects of vincristine are driven by neuroinflammatory processes, including the activation of the Nod-like receptor 3 (NLRP3) inflammasome. In this study, we aimed to test the effects of MCC950, a specific NLRP3 inhibitor, on the prevention of vincristine-induced adverse effects as well as tumor progression and vincristine efficacy in NOD/SCID/interleukin-2 receptor γ-negative mice patient-derived xenografts of ALL. We demonstrate that co-administration of MCC950 effectively prevented the development of mechanical allodynia, motor impairment, and weight loss and significantly improved the overall well-being of the animals without negatively impacting the in vivo efficacy of vincristine as a single agent or in combination with standard-of-care drugs. These results provide proof of principle that the adverse effects of vincristine chemotherapy can be prevented using NLRP3 inflammasome inhibitors and provide new options for the development of effective treatment strategies
Towards Consumption-Based Carbon Inequality Metrics: Socioeconomic and Demographic Insights from Chinese Households
Full text linkThe choice of carbon inequality metrics can significantly influence demand-side mitigation policies and their equity outcomes. We propose integrated carbon inequality metrics, including juxtaposing carbon inequality with economic inequality, disparity ratios across income and age groups, and structural income–urbanization inequality patterns. We then apply these new metrics and use the household expenditure survey data from China Family Panel Studies as a case study to examine household consumption-based carbon emissions in China. We assess the extent to which household consumption patterns, household expenditure, age, and urbanization contribute to the gap in per-capita household carbon footprints (CF) across income groups. We find that in relative terms, the top 20% income group accounts for 38% of total emissions, whereas the bottom 20% emit about 8% in China. Per-capita CFs vary slightly widely in their inequality than expenditure. The CF disparity ratios of all eight consumption categories across provinces concentrate around 4.5. CF disparity ratios of households with elderly members range from 1 to 3 and decrease with increasing household size. Rural CF-Gini exhibit a slightly wider range (0.15 to 0.52) than urban CF-Gini (0.16 to 0.42). Per capita CF of urban inhabitants was substantially larger than that of the rural ones, with 8.83 tCO2 per capita in urban regions vs. 2.68 tCO2 in rural regions. This study provides a nuanced understanding of within-country disparities to inform equitable demand-side mitigation solutions
Flash drought prediction using deep learning
Full text linkFlash droughts are rapid, short-term drought events that develop within weeks, driven by factors such as low rainfall, high temperatures, and strong winds, which deplete soil moisture and stress vegetation. These events have profound agricultural, economic, and ecological impacts, yet the use of machine learning to predict flash droughts remains underexplored, hindered by challenges like imbalanced datasets and limited data. This study addresses these issues by applying Convolutional neural networks (CNNs) to predict flash droughts in Eastern Australia, a region prone to such events. We identified flash droughts from 2001 to 2022, training the model with data from 2001–2015, validating it on 2016–2017 data, and testing it on 2018–2022 data. The model’s performance was evaluated across drought duration, spatial distribution, and seasonal variability. Achieving a balanced accuracy of 80% and an Area under the curve of 93%, the CNN demonstrated strong predictive capability. However, it tended to overestimate the spatial extent of droughts, indicating areas for future improvement. These results highlight the potential of deep learning in flash drought prediction, offering valuable insights for early warning systems and drought management strategies
An Integrated, 3D Conformal Electronic Device for Bioimpedance Sensing in Surgical Endoscopy
No full textMinimally invasive surgeries (MIS), in conjunction with endoscopy, offer significant advantages over traditional methods for treating colorectal cancer, including minimal incisions, faster recovery, and improved overall efficiency. Over the years, several miniaturised sensors have been developed to enhance procedures with soft medical robots by providing proprioceptive, exteroceptive, and diagnostic insights essential for a successful operation. Among them, tactile sensors are by far the most common sensor type in the diagnosis of tissue malignancy based on the difference in mechanical properties between healthy and cancerous tissues. However, such difference is generally nuanced until a late-stage malignancy. In contrast, sensing through tissue electrical properties, such as bioelectrical impedance, offers an alternative solution for distinguishing abnormalities due to the pronounced differences between various tissue stages. However, the integration of impedance sensors onto soft medical robots has not been fully explored. Onsite impedance sensing necessitates a low interfacial impedance at the sensor-tissue interface, whilst maintaining intimate contact with bio-tissue. This thesis proposes a 3D curvilinear electronic device based on mesoporous gold (mAu) sensors integrated onto an endoscope, offering significantly lower interfacial impedance, high sensitivity compared to conventional flat gold, and conformal contact with the bio-tissue during onsite sensing procedures. Employing a combination of top-down lithography, bottom-up electrochemical deposition, and a liquid-based transfer technique, the thesis project successfully demonstrated the integration of mesoporous electrodes onto the curvilinear surface of an endoscopic-based soft robotic end-effector. The results showed that mAu electrodes exhibit interfacial impedance an order of magnitude lower than flat Au in biomimetic fluids and approximately 2.5 times higher bioimpedance sensitivity. These findings highlight their promising potential for bioimpedance-based sensing applications
Implementing Circular Economy Strategies in Prefabricated Timber Construction in Response to Climate Change
Full text linkThe Circular Economy (CE) model facilitates the construction sectors’ capability to
achieve Sustainable Development Goals (SDGs). Promoting resource efficiency in
timber construction while minimising timber waste are the primary goals of the Circular
Economy (CE) model. The reuse of timber materials has been proposed as the primary
strategy for closing the material loop, and the reuse of prefabricated construction
components has emerged as a potential solution to managing construction waste. This
study aims to develop a decision-making framework to evaluate different reuse options
for prefabricated timber buildings after their end-of-use (EOU) and develop artificial
intelligence (AI)-powered CE system functionalities for promoting sustainable timber
buildings. To this end, the study identifies barriers to reusing timber materials and
components of prefabricated timber buildings after their EOU. It addresses the various
obstacles to timber reuse, namely, economic, technical, and social challenges. The most
salient challenges identified by this study include the cost of design for reuse (DfR),
technical issues involved in the disassembly of the timber structures, design and
compatibility concerns related to timber reusing purposes, the lack of knowledge and
specialised skills for disassembly and reuse of timber structures and products, and finally,
the negative cultural perception associated with reused timber products.
This research seeks to identify the most viable reuse option among alternatives for
prefabricated timber buildings after their EOU. A decision-making framework is
developed to evaluate different reuse options in place of the refabrication option,
including biocycle 1, biocycle 2, elemental retrofit, transformation, and adaptive reuse.
The most viable reuse option was determined with the Simultaneous Evaluation of
Criteria and Alternatives (SECA) method as a multi-criteria decision-making (MCDM)
approach. The SECA method allows for simultaneous evaluation of multiple decision
criteria and alternatives, enhancing decision-making accuracy. The method identified
adaptive reuse as the most viable reuse option among alternatives, followed by biocycle
2 and transformation. These results were validated with the Technique for Order
Preference by Similarity to Ideal Solution (TOPSIS) method. Adaptive reuse in this
context refers to the process of repurposing the function of a prefabricated timber
building after its EOU into a vibrant and multifunctional space. This option emphasises a
renovation strategy that retains and enhances the existing timber structure, preserving
its historic character. Besides, key elements such as timber walls and trusses are
exposed and reinforced. Finally, the study presents the potential benefits, functions, and challenges of
implementing AI in timber construction to promote CE purposes. It focuses on an AI
powered CE system in timber construction to examine the implications of integrating AI
into the CE model. An outline is presented of the key AI functions which improve the
efficiency of the proposed AI-powered system. These functions fall into the categories of
circular design optimisation, material management, and real-time monitoring and
assessment of building performance groups. Additionally, this study demonstrates three
potential benefits of integrating AI and CE in timber construction are: environmental and
resource efficiency, operational and project efficiency, and economic and business
innovation. This section concludes with an examination of the potential challenges of
implementing AI in timber construction to promote CE purposes, challenges which relate
to the domains of data and technological integration, as well as issues pertaining to
finance and resources, organisation and industry.
The novel contribution of this research is the empirical evidence it offers in the form of
combined qualitative and quantitative data, along with practical recommendations for
enhancing the adoption of an AI-powered CE system to improve sustainability in the
timber construction sector. By providing policymakers and industry practitioners with
actionable insight such as key decision criteria and barriers to timber reuse, this study
facilitates the development of strategies and policies for promoting long-term
sustainability in construction practices
Uncovering the antifungal potential of Cannabidiol and Cannabidivarin
No full textFungal infections pose a major threat to human health with increasing incidence of antifungal resistance globally. Despite the need for novel antifungal drugs, few are currently in clinical development. Here we evaluate the antifungal activity of five phytocannabinoids against several clinically relevant fungal pathogens, with a focus on the priority pathogen Cryptococcus neoformans. Our results demonstrate that Cannabidiol (CBD), and particularly Cannabidivarin (CBDV), have broad activity against C. neoformans and other fungal pathogens, including dermatophytes that cause common tinea. We found that both CBD and CBDV acted in a fungicidal man-ner and prevented biofilm formation in C. neoformans. Phytocannabinoid treatment impeded factors important for virulence and antifungal resistance, including reduced capsule size and disruption of mature biofilms. Proteomics analysis revealed that the antifungal activity of CBD and CBDV was linked to destabilisation of the membrane, alterations in ergosterol biosynthesis, disruption of metabolic pathways, as well as selective involvement of mitochondrial-associated proteins. We next tested the ability of CBD to topically clear a C. neoformans fungal infection in vivo using the Galleria mellonella burn wound model, and we observed greatly improved survival in the CBD treated larvae. This study illustrates the potential of phytocannabinoids as antifungal treatments and opens up new routes towards development of novel antifungal drugs