Artificial intelligence for geoscience: Progress, challenges, and perspectives

This paper explores the evolution of geoscientific inquiry, tracing the progression from traditional physics-based models to modern data-driven approaches facilitated by significant advancements in artificial intelligence (AI) and data collection techniques. Traditional models, which are grounded in physical and numerical frameworks, provide robust explanations by explicitly reconstructing underlying physical processes. However, their limitations in comprehensively capturing Earth's complexities and uncertainties pose challenges in optimization and real-world applicability. In contrast, contemporary data-driven models, particularly those utilizing machine learning (ML) and deep learning (DL), leverage extensive geoscience data to glean insights without requiring exhaustive theoretical knowledge. ML techniques have shown promise in addressing Earth science-related questions. Nevertheless, challenges such as data scarcity, computational demands, data privacy concerns, and the "black-box" nature of AI models hinder their seamless integration into geoscience. The integration of physics-based and data-driven methodologies into hybrid models presents an alternative paradigm. These models, which incorporate domain knowledge to guide AI methodologies, demonstrate enhanced efficiency and performance with reduced training data requirements. This review provides a comprehensive overview of geoscientific research paradigms, emphasizing untapped opportunities at the intersection of advanced AI techniques and geoscience. It examines major methodologies, showcases advances in large-scale models, and discusses the challenges and prospects that will shape the future landscape of AI in geoscience. The paper outlines a dynamic field ripe with possibilities, poised to unlock new understandings of Earth's complexities and further advance geoscience exploration.

Pollen viability, longevity, and function in angiosperms: key drivers and prospects for improvement

Pollen grains are central to sexual plant reproduction and their viability and longevity/storage are critical for plant physiology, ecology, plant breeding, and many plant product industries. Our goal is to present progress in assessing pollen viability/longevity along with recent advances in our understanding of the intrinsic and environmental factors that determine pollen performance: the capacity of the pollen grain to be stored, germinate, produce a pollen tube, and fertilize the ovule. We review current methods to measure pollen viability, with an eye toward advancing basic research and biotechnological applications. Importantly, we review recent advances in our understanding of how basic aspects of pollen/stigma development, pollen molecular composition, and intra- and intercellular signaling systems interact with the environment to determine pollen performance. Our goal is to point to key questions for future research, especially given that climate change will directly impact pollen viability/longevity. We find that the viability and longevity of pollen are highly sensitive to environmental conditions that affect complex interactions between maternal and paternal tissues and internal pollen physiological events. As pollen viability and longevity are critical factors for food security and adaptation to climate change, we highlight the need to develop further basic research for better understanding the complex molecular mechanisms that modulate pollen viability and applied research on developing new methods to maintain or improve pollen viability and longevity.

A user-friendly method to get automated pollen analysis from environmental samples

Automated pollen analysis is not yet efficient on environmental samples containing many pollen taxa and debris, which are typical in most pollen-based studies. Contrary to classification, detection remains overlooked although it is the first step from which errors can propagate. Here, we investigated a simple but efficient method to automate pollen detection for environmental samples, optimizing workload and performance. We applied the YOLOv5 algorithm on samples containing debris and c. 40 Mediterranean plant taxa, designed and tested several strategies for annotation, and analyzed variation in detection errors. About 5% of pollen grains were left undetected, while 5% of debris were falsely detected as pollen. Undetected pollen was mainly in poor-quality images, or of rare and irregular morphology. Pollen detection remained effective when applied to samples never seen by the algorithm, and was not improved by spending time to provide taxonomic details. Pollen detection of a single model taxon reduced annotation workload, but was only efficient for morphologically differentiated taxa. We offer guidelines to plant scientists to analyze automatically any pollen sample, providing sound criteria to apply for detection while using common and user-friendly tools. Our method contributes to enhance the efficiency and replicability of pollen-based studies.

Hot topics in allergen immunotherapy, 2023: Current status and future perspective

The importance of allergen immunotherapy (AIT) is multifaceted, encompassing both clinical and quality-of-life improvements and cost-effectiveness in the long term. Key mechanisms of allergen tolerance induced by AIT include changes in memory type allergen-specific T- and B-cell responses towards a regulatory phenotype with decreased Type 2 responses, suppression of allergen-specific IgE and increased IgG1 and IgG4, decreased mast cell and eosinophil numbers in allergic tissues and increased activation thresholds. The potential of novel patient enrolment strategies for AIT is taking into account recent advances in biomarkers discoveries, molecular allergy diagnostics and mobile health applications contributing to a personalized approach enhancement that can increase AIT efficacy and compliance. Artificial intelligence can help manage and interpret complex and heterogeneous data, including big data from omics and non-omics research, potentially predict disease subtypes, identify biomarkers and monitor patient responses to AIT. Novel AIT preparations, such as synthetic compounds, innovative carrier systems and adjuvants, are also of great promise. Advances in clinical trial models, including adaptive, complex and hybrid designs as well as real-world evidence, allow more flexibility and cost reduction. The analyses of AIT cost-effectiveness show a clear long-term advantage compared to pharmacotherapy. Important research questions, such as defining clinical endpoints, biomarkers of patient selection and efficacy, mechanisms and the modulation of the placebo effect and alternatives to conventional field trials, including allergen exposure chamber studies are still to be elucidated. This review demonstrates that AIT is still in its growth phase and shows immense development prospects.

Multi-Element Profile Characterization of Monofloral and Polyfloral Honey from Latvia

Honey is of scientific interest mainly due to its health-promoting and antibacterial properties, which are also associated with its floral origins. However, the methods for confirming honey floral origins are quite limited and require improvements. One method suggested in the search for a multi-method approach to evaluating the floral origins of Latvian honey is inductively coupled plasma mass spectrometry (ICP-MS). This study investigated the multi-element profile of 83 honey samples of well-specified floral origins. The main findings included using Ba, Ca, Cs, Fe, and Rb as indicator elements for heather honey. The chemometric evaluation supported the use of ICP-MS for distinguishing heather honey from other types of honey. The Latvian polyfloral honey multi-element profile was defined and compared to honey samples with other geographical origins. Additionally, the multi-element profiles of buckwheat, clover, and polyfloral honey proteins were investigated to clarify whether the majority of elements were bound with proteins or not. Preliminary results indicated that Ca, K, Mg, Mn, Na, and Sr were mainly found in non-protein-bound forms, while the majority of Al, Cu, Ni, and Zn were in the form of large chemical structures (>10 kDa).

Deductive automated pollen classification in environmental samples via exploratory deep learning and imaging flow cytometry

Pollen and tracheophyte spores are ubiquitous environmental indicators at local and global scales. Palynology is typically performed manually by microscopic analysis; a specialised and time-consuming task limited in taxonomical precision and sampling frequency, therefore restricting data quality used to inform climate change and pollen forecasting models. We build on the growing work using AI (artificial intelligence) for automated pollen classification to design a flexible network that can deal with the uncertainty of broad-scale environmental applications. We combined imaging flow cytometry with Guided Deep Learning to identify and accurately categorise pollen in environmental samples; here, pollen grains captured within c. 5500 Cal yr BP old lake sediments. Our network discriminates not only pollen included in training libraries to the species level but, depending on the sample, can classify previously unseen pollen to the likely phylogenetic order, family and even genus. Our approach offers valuable insights into the development of a widely transferable, rapid and accurate exploratory tool for pollen classification in 'real-world' environmental samples with improved accuracy over pure deep learning techniques. This work has the potential to revolutionise many aspects of palynology, allowing a more detailed spatial and temporal understanding of pollen in the environment with improved taxonomical resolution.

Bumblebee flower constancy and pollen diversity over time

Bees often focus their foraging effort on a few or even a single flower species, even if other equally rewarding flower species are present. Although this phenomenon-called flower constancy-has been widely documented during single foraging trips, it is largely unknown if the behavior persists over longer time periods, especially under field conditions with large temporal variations of resources. We studied the pollen diet of individuals from nine different Bombus terrestris colonies for up to 6 weeks, to investigate flower constancy and pollen diversity of individuals and colonies, and how these change over time. We expected high degrees of flower constancy and foraging consistency over time, based on foraging theory and previous studies. Instead, we found that only 23% of the pollen foraging trips were flower constant. The fraction of constant pollen samples did not change over the study period, although repeatedly sampled individuals that were flower constant once often showed different preferences at other sampling occasions. The similarity of pollen composition in samples collected by the same individuals at different occasions dropped with time. This suggests that the flower preferences change in response to shifting floral resources. The average diversity of pollen from single foraging trips was around 2.5 pollen types, while the colony-level pollen diversity was about three times higher. How rapidly preferences change in response to shifting resources, and if this differs between and within bee species depending on factors such as size, should be the focus of future research.

EAACI guidelines on environmental science in allergic diseases and asthma - Leveraging artificial intelligence and machine learning to develop a causality model in exposomics

Allergic diseases and asthma are intrinsically linked to the environment we live in and to patterns of exposure. The integrated approach to understanding the effects of exposures on the immune system includes the ongoing collection of large-scale and complex data. This requires sophisticated methods to take full advantage of what this data can offer. Here we discuss the progress and further promise of applying artificial intelligence and machine-learning approaches to help unlock the power of complex environmental data sets toward providing causality models of exposure and intervention. We discuss a range of relevant machine-learning paradigms and models including the way such models are trained and validated together with examples of machine learning applied to allergic disease in the context of specific environmental exposures as well as attempts to tie these environmental data streams to the full representative exposome. We also discuss the promise of artificial intelligence in personalized medicine and the methodological approaches to healthcare with the final AI to improve public health.

Exploring the Intersection of Artificial Intelligence and Clinical Healthcare: A Multidisciplinary Review

Artificial intelligence (AI) plays a more and more important role in our everyday life due to the advantages that it brings when used, such as 24/7 availability, a very low percentage of errors, ability to provide real time insights, or performing a fast analysis. AI is increasingly being used in clinical medical and dental healthcare analyses, with valuable applications, which include disease diagnosis, risk assessment, treatment planning, and drug discovery. This paper presents a narrative literature review of AI use in healthcare from a multi-disciplinary perspective, specifically in the cardiology, allergology, endocrinology, and dental fields. The paper highlights data from recent research and development efforts in AI for healthcare, as well as challenges and limitations associated with AI implementation, such as data privacy and security considerations, along with ethical and legal concerns. The regulation of responsible design, development, and use of AI in healthcare is still in early stages due to the rapid evolution of the field. However, it is our duty to carefully consider the ethical implications of implementing AI and to respond appropriately. With the potential to reshape healthcare delivery and enhance patient outcomes, AI systems continue to reveal their capabilities.

An Outlook for Deep Learning in Ecosystem Science

Rapid advances in hardware and software, accompanied by public- and private-sector investment, have led to a new generation of data-driven computational tools. Recently, there has been a particular focus on deep learning—a class of machine learning algorithms that uses deep neural networks to identify patterns in large and heterogeneous datasets. These developments have been accompanied by both hype and scepticism by ecologists and others. This review describes the context in which deep learning methods have emerged, the deep learning methods most relevant to ecosystem ecologists, and some of the problem domains they have been applied to. Deep learning methods have high predictive performance in a range of ecological contexts, leveraging the large data resources now available. Furthermore, deep learning tools offer ecosystem ecologists new ways to learn about ecosystem dynamics. In particular, recent advances in interpretable machine learning and in developing hybrid approaches combining deep learning and mechanistic models provide a bridge between pure prediction and causal explanation. We conclude by looking at the opportunities that deep learning tools offer ecosystem ecologists and assess the challenges in interpretability that deep learning applications pose.

DNA metabarcoding using nrITS2 provides highly qualitative and quantitative results for airborne pollen monitoring

Airborne pollen monitoring is of global socio-economic importance as it provides information on presence and prevalence of allergenic pollen in ambient air. Traditionally, this task has been performed by microscopic investigation, but novel techniques are being developed to automate this process. Among these, DNA metabarcoding has the highest potential of increasing the taxonomic resolution, but uncertainty exists about whether the results can be used to quantify pollen abundance. In this study, it is shown that DNA metabarcoding using trnL and nrITS2 provides highly improved taxonomic resolution for pollen from aerobiological samples from the Netherlands. A total of 168 species from 143 genera and 56 plant families were detected, while using a microscope only 23 genera and 22 plant families were identified. NrITS2 produced almost double the number of OTUs and a much higher percentage of identifications to species level (80.1%) than trnL (27.6%). Furthermore, regressing relative read abundances against the relative abundances of microscopically obtained pollen concentrations showed a better correlation for nrITS2 (R2 = 0.821) than for trnL (R2 = 0.620). Using three target taxa commonly encountered in early spring and fall in the Netherlands (Alnus sp., Cupressaceae/Taxaceae and Urticaceae) the nrITS2 results showed that all three taxa were dominated by one or two species (Alnus glutinosa/incana, Taxus baccata and Urtica dioica). Highly allergenic as well as artificial hybrid species were found using nrITS2 that could not be identified using trnL or microscopic investigation (Alnus × spaethii, Cupressus arizonica, Parietaria spp.). Furthermore, perMANOVA analysis indicated spatiotemporal patterns in airborne pollen trends that could be more clearly distinguished for all taxa using nrITS2 rather than trnL. All results indicate that nrITS2 should be the preferred marker of choice for molecular airborne pollen monitoring.

Segmenting 20 Types of Pollen Grains for the Cretan Pollen Dataset v1 (CPD-1)

Pollen analysis and the classification of several pollen species is an important task in melissopalynology. The development of machine learning or deep learning based classification models depends on available datasets of pollen grains from various plant species from around the globe. In this paper, Cretan Pollen Dataset v1 (CPD-1) is presented, which is a novel dataset of grains from 20 pollen species from plants gathered in Crete, Greece. The pollen grains were prepared and stained with fuchsin, in order to be captured by a camera attached to a microscope under a ×400 magnification. In addition, a pollen grain segmentation method is presented, which segments and crops each unique pollen grain and achieved an overall detection accuracy of 92%. The final dataset comprises 4034 segmented pollen grains of 20 different pollen species, as well as the raw data and ground truth, as annotated by an expert. The developed dataset is publicly accessible, which we hope will accelerate research in melissopalynology.