Explore the vast range of titles available.

Natural hazards have the potential to cause catastrophic damage and significant socioeconomic loss. The actual damage and loss observed in the recent decades has shown an increasing trend. As a result, disaster managers need to take a growing responsibility to proactively protect their communities by developing efficient management strategies. A number of research studies apply artificial intelligence (AI) techniques to process disaster-related data for supporting informed disaster management. This study provides an overview of current applications of AI in disaster management during its four phases: mitigation, preparedness, response, and recovery. It presents example applications of different AI techniques and their benefits for supporting disaster management at different phases, as well as some practical AI-based decision support tools. We find that the majority of AI applications focus on the disaster response phase. This study also identifies challenges to inspire the professional community to advance AI techniques for addressing them in future research.

Doc number: 15 Abstract: The production of cellulosic ethanol from biomass is considered a promising alternative to reliance on diminishing supplies of fossil fuels, providing a sustainable option for fuels production in an environmentally compatible manner. The conversion of lignocellulosic biomass to biofuels through a biological route usually suffers from the intrinsic recalcitrance of biomass owing to the complicated structure of plant cell walls. Currently, a pretreatment step that can effectively reduce biomass recalcitrance is generally required to make the polysaccharide fractions locked in the intricacy of plant cell walls to become more accessible and amenable to enzymatic hydrolysis. Dilute acid and hydrothermal pretreatments are attractive and among the most promising pretreatment technologies that enhance sugar release performance. This review highlights our recent understanding on molecular structure basis for recalcitrance, with emphasis on structural transformation of major biomass biopolymers (i.e., cellulose, hemicellulose, and lignin) related to the reduction of recalcitrance during dilute acid and hydrothermal pretreatments. The effects of these two pretreatments on biomass porosity as well as its contribution on reduced recalcitrance are also discussed.

Traditional biomedical artificial intelligence (AI) models, designed for specific tasks or modalities, often exhibit limited flexibility in real-world deployment and struggle to utilize holistic information. Generalist AI holds the potential to address these limitations due to its versatility in interpreting different data types and generating tailored outputs for diverse needs. However, existing biomedical generalist AI solutions are typically heavyweight and closed source to researchers, practitioners and patients. Here, we describe BiomedGPT, the first open-source and lightweight vision–language foundation model, designed as a generalist capable of performing various biomedical tasks. BiomedGPT achieved state-of-the-art results in 16 out of 25 experiments while maintaining a computing-friendly model scale. We also conducted human evaluations to assess the capabilities of BiomedGPT in radiology visual question answering, report generation and summarization. BiomedGPT exhibits robust prediction ability with a low error rate of 3.8% in question answering, satisfactory performance with an error rate of 8.3% in writing complex radiology reports, and competitive summarization ability with a nearly equivalent preference score to human experts. Our method demonstrates that effective training with diverse data can lead to more practical biomedical AI for improving diagnosis and workflow efficiency. An open-source and computing-friendly vision–language model achieves state-of-the-art accuracy in 16 out of 25 biomedical tasks, with promising performance in a series of potential clinical applications.

Lignin is a chemically complex polymer that lends woody plants and trees their rigidity. Humans have traditionally either left it intact to lend rigidity to their own wooden constructs, or burned it to generate heat and sometimes power. With the advent of major biorefining operations to convert cellulosic biomass into ethanol and other liquid fuels, researchers are now exploring how to transform the associated leftover lignin into more diverse and valuable products. Ragauskas et al. ( 10.1126/science.1246843 ) review recent developments in this area, ranging from genetic engineering approaches that tune lignin properties at the source, to chemical processing techniques directed toward extracting lignin in the biorefinery and transforming it into high-performance plastics and a variety of bulk and fine chemicals. Research and development activities directed toward commercial production of cellulosic ethanol have created the opportunity to dramatically increase the transformation of lignin to value-added products. Here, we highlight recent advances in this lignin valorization effort. Discovery of genetic variants in native populations of bioenergy crops and direct manipulation of biosynthesis pathways have produced lignin feedstocks with favorable properties for recovery and downstream conversion. Advances in analytical chemistry and computational modeling detail the structure of the modified lignin and direct bioengineering strategies for future targeted properties. Refinement of biomass pretreatment technologies has further facilitated lignin recovery, and this coupled with genetic engineering will enable new uses for this biopolymer, including low-cost carbon fibers, engineered plastics and thermoplastic elastomers, polymeric foams, fungible fuels, and commodity chemicals.

Interactions of water with cellulose are of both fundamental and technological importance. Here, we characterize the properties of water associated with cellulose using deuterium labeling, neutron scattering and molecular dynamics simulation. Quasi-elastic neutron scattering provided quantitative details about the dynamical relaxation processes that occur and was supported by structural characterization using small-angle neutron scattering and X-ray diffraction. We can unambiguously detect two populations of water associated with cellulose. The first is “non-freezing bound” water that gradually becomes mobile with increasing temperature and can be related to surface water. The second population is consistent with confined water that abruptly becomes mobile at ~260 K, and can be attributed to water that accumulates in the narrow spaces between the microfibrils. Quantitative analysis of the QENS data showed that, at 250 K, the water diffusion coefficient was 0.85 ± 0.04 × 10 −10  m 2 sec −1 and increased to 1.77 ± 0.09 × 10 −10  m 2 sec −1 at 265 K. MD simulations are in excellent agreement with the experiments and support the interpretation that water associated with cellulose exists in two dynamical populations. Our results provide clarity to previous work investigating the states of bound water and provide a new approach for probing water interactions with lignocellulose materials.

Twitter is a popular source for the monitoring of healthcare information and public disease. However, there exists much noise in the tweets. Even though appropriate keywords appear in the tweets, they do not guarantee the identification of a truly health-related tweet. Thus, the traditional keyword-based classification task is largely ineffective. Algorithms for word embeddings have proved to be useful in many natural language processing (NLP) tasks. We introduce two algorithms based on an existing word embedding learning algorithm: the continuous bag-of-words model (CBOW). We apply the proposed algorithms to the task of recognizing healthcare-related tweets. In the CBOW model, the vector representation of words is learned from their contexts. To simplify the computation, the context is represented by an average of all words inside the context window. However, not all words in the context window contribute equally to the prediction of the target word. Greedily incorporating all the words in the context window will largely limit the contribution of the useful semantic words and bring noisy or irrelevant words into the learning process, while existing word embedding algorithms also try to learn a weighted CBOW model. Their weights are based on existing pre-defined syntactic rules while ignoring the task of the learned embedding. We propose learning weights based on the words’ relative importance in the classification task. Our intuition is that such learned weights place more emphasis on words that have comparatively more to contribute to the later task. We evaluate the embeddings learned from our algorithms on two healthcare-related datasets. The experimental results demonstrate that embeddings learned from the proposed algorithms outperform existing techniques by a relative accuracy improvement of over 9%.

Manually labeling data for training machine learning models is time-consuming and expensive. Therefore, it is often necessary to apply models built in one domain to a new domain. However, existing approaches do not evaluate the quality of intermediate features that are learned in the process of transferring from the source domain to the target domain, which results in the potential for sub-optimal features. Also, transfer learning models in existing work do not provide optimal results for a new domain. In this paper, we first propose a fast subspace sampling demons (SSD) method to learn intermediate subspace features from two domains and then evaluate the quality of the learned features. To show the applicability of our model, we test our model using a synthetic dataset as well as several benchmark datasets. Extensive experiments demonstrate significant improvements in classification accuracy over the state of the art.

The primary obstacle to producing renewable fuels from lignocellulosic biomass is a plant's recalcitrance to releasing sugars bound in the cell wall. From a sample set of wood cores representing 1,100 individual undomesticated Populus trichocarpa trees, 47 extreme phenotypes were selected across measured lignin content and ratio of syringyl and guaiacyl units (S/G ratio). This subset was tested for total sugar release through enzymatic hydrolysis alone as well as through combined hot-water pretreatment and enzymatic hydrolysis using a high-throughput screening method. The total amount of glucan and xylan released varied widely among samples, with total sugar yields of up to 92% of the theoretical maximum. A strong negative correlation between sugar release and lignin content was only found for pretreated samples with an S/G ratio < 2.0. For higher S/G ratios, sugar release was generally higher, and the negative influence of lignin was less pronounced. When examined separately, only glucose release was correlated with lignin content and S/G ratio in this manner, whereas xylose release depended on the S/G ratio alone. For enzymatic hydrolysis without pretreatment, sugar release increased significantly with decreasing lignin content below 20%, irrespective of the S/G ratio. Furthermore, certain samples featuring average lignin content and S/G ratios exhibited exceptional sugar release. These facts suggest that factors beyond lignin and S/G ratio influence recalcitrance to sugar release and point to a critical need for deeper understanding of cell-wall structure before plants can be rationally engineered for reduced recalcitrance and efficient biofuels production.

Influenza, categorized as one of the emergent infectious diseases, presents a substantial public health concern due to its capacity to trigger extensive epidemics and global pandemics. Every recent pandemic of human influenza has been attributed to avian influenza viruses (AIVs), underscoring their pandemic potential and the associated public health risks. Interestingly, there remains a significant knowledge gap concerning the mechanisms that sustain the survival and proliferation of these viruses within their natural avian reservoirs. Migratory waterfowl, in particular mallard ( ), plays a crucial role as potential reservoirs, facilitating the spread of the virus through their migratory patterns. In order to better understand the factors that contribute to AIV spillover and hotspots, we built a mechanistic transmission model and investigated the likelihood of an AIV hotspot for the animal-to-human spillover of the virus in a region in Kansas. Our findings challenge the notion that the size of the overall mallard population is a reliable predictor of spillover hazard. To study this aspect, we systematically evaluated weekly trends in both the total population density and the total infected population density. In more than 70% of the locations studied, these two indicators showed periods with opposite trends. This conclusion stresses the importance of developing and calibrating compartmental models that can capture the diffusion of the virus within the reservoir and its spatiotemporal distribution due to animal movement.

Enthusiasm for using biotech to meet societal energy challenges is at levels not seen since the early 1980s, when understanding and capability in the life sciences were at a radically different stage of development than today. For cellulosic ethanol to become a reality, biotechnological solutions should focus on optimizing the conversion of biomass to sugars.