Artificial intelligence in radiotherapy

Artificial intelligence (AI) has already been implemented widely in the medical field in the recent years. This paper first reviews the background of AI and radiotherapy. Then it explores the basic concepts of different AI algorithms and machine learning methods, such as neural networks, that are available to us today and how they are being implemented in radiotherapy and diagnostic processes, such as medical imaging, treatment planning, patient simulation, quality assurance and radiation dose delivery. It also explores the ongoing research on AI methods that are to be implemented in radiotherapy in the future. The review shows very promising progress and future for AI to be widely used in various areas of radiotherapy. However, basing on various concerns such as availability and security of using big data, and further work on polishing and testing AI algorithms, it is found that we may not ready to use AI primarily in radiotherapy at the moment.     

Current topics in artificial insemination of sheep

There have been developments in several aspects of artificial insemination (AI) in recent years, some of which have been directly responsible for proliferation of AI in the sheep-breeding industries of several countries. The most notable advances have probably been associated with the development of intrauterine insemination by laparoscopy. There is potential for refinement of some of the related techniques, particularly in the area of control of ovulation and definition of appropriate times and optimum doses of spermatozoa for insemination. It is unlikely that laparoscopic AI will be developed sufficiently that it will become readily affordable, and therefore widely practised, by commercial producers. Unfortunately, there has been little progress in the past few years in improvement of the methods of cryopreservation of ram semen. There is considerable potential for AI to have a significant impact on the genetic improvement of sheep, though this has yet to be evaluated in practice. However, if the full potential of AI in sheep is to be realized, it will likely only happen when methods of freezing semen are improved sufficiently that cervical or even vaginal insemination can be widely used with frozen-thawed semen, or when practicable methods of deep cervical or intrauterine insemination through the cervix are developed.     

AI in 3D compound design

The success of Artificial Intelligence (AI) across a wide range of domains has fuelled significant interest in its application to designing novel compounds and screening compounds against a specific target. However, many existing AI methods either do not account for the 3D structure of the target at all or struggle to capture meaningful spatial information from the target. In this Opinion, we highlight a range of recent structure-aware approaches which utilise deep learning for compound design and virtual screening. We discuss how such methods can be better integrated into existing drug discovery pipelines by facilitating the design of compounds which conform to a specified design hypothesis and by uncovering key protein-ligand interactions which can be used to aid molecule design.     

Artificial intelligence and machine learning for medical imaging: A technology review

Artificial intelligence (AI) has recently become a very popular buzzword, as a consequence of disruptive technical advances and impressive experimental results, notably in the field of image analysis and processing. In medicine, specialties where images are central, like radiology, pathology or oncology, have seized the opportunity and considerable efforts in research and development have been deployed to transfer the potential of AI to clinical applications. With AI becoming a more mainstream tool for typical medical imaging analysis tasks, such as diagnosis, segmentation, or classification, the key for a safe and efficient use of clinical AI applications relies, in part, on informed practitioners. The aim of this review is to present the basic technological pillars of AI, together with the state-of-the-art machine learning methods and their application to medical imaging. In addition, we discuss the new trends and future research directions. This will help the reader to understand how AI methods are now becoming an ubiquitous tool in any medical image analysis workflow and pave the way for the clinical implementation of AI-based solutions.     

Science-policy interfaces for biodiversity: dynamic learning environments for successful impact

To address the pressing problems associated with biodiversity loss, changes in awareness and behaviour are required from decision makers in all sectors. Science-policy interfaces (SPIs) have the potential to play an important role, and to achieve this effectively, there is a need to understand better the ways in which existing SPIs strive for effective communication, learning and behavioural change. Using a series of test cases across the world, we assess a range of features influencing the effectiveness of SPIs through communication and argumentation processes, engagement of actors and other aspects that contribute to potential success. Our results demonstrate the importance of dynamic and iterative processes of interaction to support effective SPI work. We stress the importance of seeing SPIs as dynamic learning environments and we provide recommendations for how they can enhance success in meeting their targeted outcomes. In particular, we recommend building long-term trust, creating learning environments, fostering participation and ownership of the process and building capacity to combat silo thinking. Processes to enable these changes may include, for example, inviting and integrating feedback, extended peer review and attention to contextualising knowledge for different audiences, and time and sustained effort dedicated to trust-building and developing common languages. However there are no ‘one size fits all’ solutions, and methods must be adapted to context and participants. Creating and maintaining effective dynamic learning environments will both require and encourage changes in institutional and individual behaviours: a challenging agenda, but one with potential for positive feedbacks to maintain momentum.     

Open data and algorithms for open science in AI-driven molecular informatics

Recent years have seen a sharp increase in the development of deep learning and artificial intelligence-based molecular informatics. There has been a growing interest in applying deep learning to several subfields, including the digital transformation of synthetic chemistry, extraction of chemical information from the scientific literature, and AI in natural product-based drug discovery. The application of AI to molecular informatics is still constrained by the fact that most of the data used for training and testing deep learning models are not available as FAIR and open data. As open science practices continue to grow in popularity, initiatives which support FAIR and open data as well as open-source software have emerged. It is becoming increasingly important for researchers in the field of molecular informatics to embrace open science and to submit data and software in open repositories. With the advent of open-source deep learning frameworks and cloud computing platforms, academic researchers are now able to deploy and test their own deep learning models with ease. With the development of new and faster hardware for deep learning and the increasing number of initiatives towards digital research data management infrastructures, as well as a culture promoting open data, open source, and open science, AI-driven molecular informatics will continue to grow. This review examines the current state of open data and open algorithms in molecular informatics, as well as ways in which they could be improved in future.     

A technique to artificially inseminate leafcutter ants

Artificial insemination (AI), the instrumental transfer of semen from the male to female reproductive organs, offers excellent opportunities to study mating system adaptations as it allows paternity to be experimentally manipulated. AI techniques have been developed for many animals, but rarely for ants, where they would be particularly useful as most species do not mate under controlled lab conditions. Here, we describe an AI technique for Atta leafcutter ants involving (1) the collection of ejaculates via induction of natural ejaculation, (2) storage in glass capillaries, and (3) transfer to queens using a modified AI equipment as used for honeybees. Queens were fixed and anesthetized in a queen holder, after which the sting chamber was opened with two steel hooks, the tip of the semen-containing glass capillary was inserted into the bursa copulatrix and the semen slowly expelled. Sperm was successfully stored in the spermatheca of queens, and some queens produced a small colony as a result. We could furthermore confirm the earlier observations that Atta semen is directly transferred to the spermatheca rather than to the bursa copulatrix as in most other eusocial insects. The technique that we present here can offer novel opportunities to study mating events such as sperm transfer, sperm competition, and cryptic female choice in ants. At present, the number of queens that produce colonies after AI remains low. However, this number will likely increase, as our results indicate that rearing conditions after AI influence colony founding success of artificially inseminated Atta queens.     

Mathematical models, explanation, laws, and evolutionary biology

It is commonly agreed in the literature on laws of nature that there are at least two necessary conditions for lawhood--that a law must have empirical content and that it must be universal. The main reason offered for the requirement that laws be empirical is as follows: a priori statements are consistent with any imaginable set of observations, so they cannot be informative about the world and therefore they cannot provide explanations. However, we care about laws because we think that laws provide explanations and allow us to make predictions. Thus, if one of the functions of laws is to provide explanations and a priori propositions cannot fulfill this function, they cannot properly be viewed as laws. In this paper, I will aim to show that this argument for the claim that laws must be empirical does not work.     

Critical comparison of consensus methods for molecular sequences.

Consensus methods are recognized as valuable tools for data analysis, especially when some sort of data aggregation is desired. Although consensus methods for sequences play a vital role in molecular biology, researchers pay little heed to the features and limitations of such methods, and so there are risks that criteria for constructing consensus sequences will be misused or misunderstood. To understand better the issues involved, we conducted a critical comparison of nine consensus methods for sequences, of which eight were used in papers appearing in this journal. We report the results of that comparison, and we make recommendations which we hope will assist researchers when they must select particular consensus methods for particular applications.     

Closing the translation gap: AI applications in digital pathology

Recent advances in artificial intelligence show tremendous promise to improve the accuracy, reproducibility, and availability of medical diagnostics across a number of medical subspecialities. This is especially true in the field of digital pathology, which has recently witnessed a surge in publications describing state-of-the-art performance for machine learning models across a wide range of diagnostic applications. Nonetheless, despite this promise, there remain significant gaps in translating applications for any of these technologies into actual clinical practice. In this review, we will first give a brief overview of the recent progress in applying AI to digitized pathology images, focusing on how these tools might be applied in clinical workflows in the near term to improve the accuracy and efficiency of pathologists. Then we define and describe in detail the various factors that need to be addressed in order to successfully close the “translation gap” for AI applications in digital pathology.     

Principles and limitations of computational microRNA gene and target finding

In 2001 there were four PubMed entries matching the word "microRNA" (miRNA). Interestingly, this number has now far exceeded 1300 and is still rapidly increasing. This more than anything demonstrates the extreme attention this field has had within a short period of time. With the large amounts of sequence data being generated, the need for analysis by computational approaches is obvious. Here, we review the general principles used in computational gene and target finding, and discuss the strengths and weaknesses of the methods. Several methods rely on detection of evolutionary conserved candidates, but recent methods have challenged this paradigm by simultaneously searching for the gene and the corresponding target(s). Whereas the early methods made predictions based on sets of hand-derived rules from precursor-miRNA structure or observed target-miRNA interactions, recent methods apply machine learning techniques. Even though these methods are already powerful, the amount of data they rely on is still limited. Since it is evident that data are continuously being generated, it must be anticipated that these methods will further improve their performance.     

The development, application and limitations of breast cancer cell lines to study tamoxifen and aromatase inhibitor resistance

Estrogen plays important roles in hormone receptor-positive breast cancer. Endocrine therapies, such as the antiestrogen tamoxifen, antagonize the binding of estrogen to estrogen receptor (ER), whereas aromatase inhibitors (AIs) directly inhibit the production of estrogen. Understanding the mechanisms of endocrine resistance and the ways in which we may better treat these types of resistance has been aided by the development of cellular models for resistant breast cancers. In this review, we will discuss what is known thus far regarding both de novo and acquired resistance to tamoxifen or AIs. Our laboratory has generated a collection of AI- and tamoxifen-resistant cell lines in order to comprehensively study the individual types of resistance mechanisms. Through the use of microarray analysis, we have determined that our cell lines resistant to a particular AI (anastrozole, letrozole, or exemestane) or tamoxifen are distinct from each other, indicating that these mechanisms can be quite complex. Furthermore, we will describe two novel de novo AI-resistant cell lines that were generated from our laboratory. Initial characterization of these cells reveals that they are distinct from our acquired AI-resistant cell models. In addition, we will review potential therapies which may be useful for overcoming resistant breast cancers through studies using endocrine resistant cell lines. Finally, we will discuss the benefits and shortcomings of cell models. Together, the information presented in this review will provide us a better understanding of acquired and de novo resistance to tamoxifen and AI therapies, the use of appropriate cell models to better study these types of breast cancer, which are valuable for identifying novel treatments and strategies for overcoming both tamoxifen and AI-resistant breast cancers.     

Interactions between parasite and vector control, animal productivity and rural welfare

These fundamental economic pressures coupled with concern for the environment and for the welfare of animals and, of course, for the welfare of humans, lead us to the fundamental conclusion that the goal of livestock production must be optimisation of inputs and outputs to satisfy the market needs. Standards will have to be set for welfare, land use patterns will have to be modified for social as well as economic reasons, and these will lead to the selection of health and production policies which optimise the long term efficiency of animal use.As we have seen, arthropods and the diseases they carry have profound impact on the efficiency of production and constrain the use of land in many areas where there are deficits of food. Consequently, we have to integrate control policies for these problems into a more orderly process of livestock development. Since we have recognised and accepted that a different series of factors intervene in different regions within countries, we must also accept that control and eradication policies will differ from area to area and from time to time. Parasite control has to be slotted into the order of priorities at different points and at different levels of intensity as the development process continues. Economic and social analysis, with the help of modelling, can provide guidance on the changes that may be needed but it is imperative that they be complemented by more effective recording and monitoring schemes. These schemes must cover animal productivity, and the full mix of factors which constrain efficient animal use. Fortunately, we now understand how to do all this, at a cost which is sustainable by any and every country.     

Artificial intelligence for carbon emissions using system of systems theory

The impact of artificial intelligence (AI) on the environment is the subject of discourse, with arguments for both positive and negative effects. There is a fine line between AI for good and AI for environmental degradation. Today, companies want to seize the benefits of AI, which distinctively involves reducing the company's carbon footprint. However, AI's carbon emissions differ as per the techniques involved in training it. As the saying goes, a coin always has two sides. Therefore, it cannot be denied that AI can be an effective tool for combating climate change, but its role in contributing to carbon emissions cannot be ignored. Multiple studies indicate that AI could be the game-changer in staving off anthropogenic climatic changes due to the deterioration of the environment and global warming. This double-edged relationship and interdependency of AI and carbon emissions are represented through a system of systems (SoS) approach. SoS states that a plan is created through multiple smaller systems, creating complexity in the design and vice versa. A complex system can be assumed as the world in general, where two individual independent systems AI and carbon emissions, when in interaction, create a complex complementary and contradictory relation, adding to the convolution of the system. This connection is demonstrated by conducting a network analysis and calculating the carbon emissions of six machine learning (ML) algorithms and deep learning (DL) models with different datasets but the same hyperparameters on a carbon emission calculator created through AI algorithms. The primary idea of this study is to encourage the AI society to create efficient AI models that may be used without compromising environmental issues. The focus should be on practicing sustainable AI, that is, sustainability from data collection to model deployment, throughout the lifecycle of AI.     

One Giant Leap for Categorizers: One Small Step for Categorization Theory

We explore humans’ rule-based category learning using analytic approaches that highlight their psychological transitions during learning. These approaches confirm that humans show qualitatively sudden psychological transitions during rule learning. These transitions contribute to the theoretical literature contrasting single vs. multiple category-learning systems, because they seem to reveal a distinctive learning process of explicit rule discovery. A complete psychology of categorization must describe this learning process, too. Yet extensive formal-modeling analyses confirm that a wide range of current (gradient-descent) models cannot reproduce these transitions, including influential rule-based models (e.g., COVIS) and exemplar models (e.g., ALCOVE). It is an important theoretical conclusion that existing models cannot explain humans’ rule-based category learning. The problem these models have is the incremental algorithm by which learning is simulated. Humans descend no gradient in rule-based tasks. Very different formal-modeling systems will be required to explain humans’ psychology in these tasks. An important next step will be to build a new generation of models that can do so.     

Exome sequencing followed by large-scale genotyping suggests a limited role for moderately rare risk factors of strong effect in schizophrenia

Schizophrenia is a severe psychiatric disorder with strong heritability and marked heterogeneity in symptoms, course, and treatment response. There is strong interest in identifying genetic risk factors that can help to elucidate the pathophysiology and that might result in the development of improved treatments. Linkage and genome-wide association studies (GWASs) suggest that the genetic basis of schizophrenia is heterogeneous. However, it remains unclear whether the underlying genetic variants are mostly moderately rare and can be identified by the genotyping of variants observed in sequenced cases in large follow-up cohorts or whether they will typically be much rarer and therefore more effectively identified by gene-based methods that seek to combine candidate variants. Here, we consider 166 persons who have schizophrenia or schizoaffective disorder and who have had either their genomes or their exomes sequenced to high coverage. From these data, we selected 5,155 variants that were further evaluated in an independent cohort of 2,617 cases and 1,800 controls. No single variant showed a study-wide significant association in the initial or follow-up cohorts. However, we identified a number of case-specific variants, some of which might be real risk factors for schizophrenia, and these can be readily interrogated in other data sets. Our results indicate that schizophrenia risk is unlikely to be predominantly influenced by variants just outside the range detectable by GWASs. Rather, multiple rarer genetic variants must contribute substantially to the predisposition to schizophrenia, suggesting that both very large sample sizes and gene-based association tests will be required for securely identifying genetic risk factors.     

Individual variation in the rate of use of tree-hole tools among wild orang-utans: implications for hominin evolution

Primate tool use varies among species, populations, and individuals. Individual variation is especially poorly understood. Orang-utans in the Sumatran swamp forest of Suaq Balimbing varied widely in rates of tool use to extract honey, ants or termites from tree holes and in the degree to which they specialized on this tree-hole tool use. We tested whether individual variation was best explained by effects of social dominance, habitat differences, or by opportunities for socially learning the skills during ontogeny. There was no evidence for the first two hypotheses. However, we found a strong relationship between tool use specialization and mean female party size, which was used as a proxy for the opportunities for socially mediated learning in a foraging context during their development. This use was justified because females are rather philopatric and their mean party size remained stable over time, thus reflecting long-term tendencies. The correlation was not an artifact of a direct effect of party size on tool use tendencies, and did not hold for males, the dispersing sex. Thus, variation in the number of opportunities for social learning explains tool use variation within populations, corroborating hypotheses for between-population variation. The emergence of human culture was accompanied by vastly improved mechanisms of social learning. In order for these improvements to be favored by natural selection, the cultural potential must have actually been expressed. Thus, a combination of strong sociability and a reliance on tool-using or other technical skills acquired through social learning must have characterized early hominins.     

The European Federation of Organisations for Medical Physics (EFOMP) White Paper: Big data and deep learning in medical imaging and in relation to medical physics profession

Big data and deep learning will profoundly change various areas of professions and research in the future. This will also happen in medicine and medical imaging in particular. As medical physicists, we should pursue beyond the concept of technical quality to extend our methodology and competence towards measuring and optimising the diagnostic value in terms of how it is connected to care outcome. Functional implementation of such methodology requires data processing utilities starting from data collection and management and culminating in the data analysis methods. Data quality control and validation are prerequisites for the deep learning application in order to provide reliable further analysis, classification, interpretation, probabilistic and predictive modelling from the vast heterogeneous big data. Challenges in practical data analytics relate to both horizontal and longitudinal analysis aspects. Quantitative aspects of data validation, quality control, physically meaningful measures, parameter connections and system modelling for the future artificial intelligence (AI) methods are positioned firmly in the field of Medical Physics profession. It is our interest to ensure that our professional education, continuous training and competence will follow this significant global development.     

Emerging deep learning methods for single-cell RNA-seq data analysis

Deep learning is making major breakthrough in several areas of bioinformatics. Anticipating that this will occur soon for the single-cell RNA-seq data analysis, we review newly published deep learning methods that help tackle computational challenges. Autoencoders are found to be the dominant approach. However, methods based on deep generative models such as generative adversarial networks (GANs) are also emerging in this area.