Search

Abstract Background and Aims Interstitial lung diseases (ILDs), encompassing both pediatric and adult cases, present a diverse spectrum of chronic conditions with variable prognosis. Despite limited therapeutic options beyond antifibrotic drugs and immunosuppressants, accurate diagnosis is challenging, often necessitating invasive procedures that may not be feasible for certain patients. Drawn against this background, experts across pediatric and adult ILD fields have joined forces in the RARE-ILD initiative to pioneer novel non-invasive diagnostic algorithms and biomarkers. Collaborating with the RARE-ILD consortium, the eurILDreg aims to comprehensively describe different ILDs, analyze genetically defined forms across age groups, create innovative diagnostic and therapeutic biomarkers, and employ artificial intelligence for data analysis. Methods The foundation of eurILDreg is built on a comprehensive parameter list developed and adopted by clinical experts, encompassing over 1,800 distinct parameters related to patient history, clinical examinations, diagnosis, lung function and biospecimen collection. This robust dataset is further enriched with daily assessments captured through the patientMpower app, including handheld spirometry and exercise tests, conducted on approximately 350 patients over the course of a year. This approach involves app-based daily assessments of quality of life, symptom tracking, handheld spirometry, saturation measurement, and the 1-min sit-to-stand test (1-STST). Additionally, pediatric data from the ChILD-EU registry will be integrated into the RARE-ILD Data Warehouse, with the ultimate goal of including a total of 4.000 ILD patients and over 100.000 biospecimen. Discussion The collaborative efforts within the consortium are poised to streamline research endeavors significantly, promising to advance patient-centered care, foster innovation, and shape the future landscape of interstitial lung disease research and healthcare practices. Trial Registration EurILDreg is registered in the German Clinical Trials Register (DRKS 00028968, 26.07.2022), and eurIPFreg is registered in ClinicalTrials.gov (NCT02951416).

Pikeperch, perch and bream are among the most traded and valued fish species in North-Eastern Europe. Therefore, it is necessary to be able to distinguish fish from different lakes and coastal sea regions to ensure a good traceability of products in the fish market and to protect both consumers and fish stocks. Untargeted metabolomics using nuclear magnetic resonance (NMR) spectroscopy is a suitable tool for this purpose. It is an established method for determining various properties of biological and living systems, such as health, origin, type, etc. Statistical methods including principal component analysis (PCA) and linear discriminant analysis (LDA) are typically applied to NMR data to correlate spectra with a particular research question.Herein we examine fish from three closely related water bodies and demonstrate that reliable determination of the water body that a particular fish originates from by traditional statistical analysis (PCA and LDA) of fish NMR spectra is not possible. In contrast, determining the fish species is possible. We proceed to show that machine learning methods perform better and that a combination of statistical analysis (LDA) and random forest (RF), a supervised machine learning technique, allows reliable determination of the originating water body, while being also tolerant to seasonal variations. This is an improvement over prior work, which has dealt with more clearly distinguished origins of fish. Exceptional accuracy was achieved in correctly assigning fish to their origin even in a scenario where two of the water bodies are connected by a river through which the fish are known to migrate. Since determining the origin of fish is important in environmental protection, we recommend following up this approach and using it as the basis of a robust tool for environmental protection and other monitoring purposes.

Abstract Prediction of chemical shift in NMR using machine learning methods is typically done with the maximum amount of data available to achieve the best results. In some cases, such large amounts of data are not available, e.g. for heteronuclei. We demonstrate a novel machine learning model that is able to achieve better results than other models for relevant datasets with comparatively low amounts of data. We show this by predicting $$^{19}F$$ 19 F and $$^{13}C$$ 13 C NMR chemical shifts of small molecules in specific solvents. Graphical Abstract

Electronic health records (EHRs) are a critical tool in healthcare and capture a wide array of patient information that can inform clinical decision-making. However, the sheer volume and complexity of EHR data present challenges for healthcare providers, particularly in fast-paced environments such as intensive care units (ICUs). To address this problem, the automatic summarization of the main problems of patients from daily progress notes can be extremely helpful. Furthermore, by accurately predicting ICU patients’ lengths of stay (LOSs), resource allocation and management can be optimized, allowing for a more efficient flow of patients within the healthcare system. This work proposes a hybrid method to summarize EHR notes and studies the potential of these summaries together with structured data for the prediction of LOSs of ICU patients. Our investigation demonstrates the effectiveness of combining extractive and abstractive summarization techniques with a concept-based method combined with a text-to-text transfer transformer (T5), which shows the most promising results. By integrating the generated summaries and diagnoses with other features, our study contributes to the accurate prediction of LOSs, with a support vector machine emerging as our best-performing classifier with an accuracy of 77.5%, surpassing existing systems and highlighting the potential for optimal allocation of resources within ICUs.

The availability of high-calorie foods with contentious nutritional content has led to a worldwide increase in chronic disease. Therefore, monitoring of eating habits and practising healthy eating habits is recommended. Clinical diet assessment methods and mobile calorie tracking apps can be used to record daily food consumption but are often not user-friendly. Convenient image-based assessment models are currently available to recognise and estimate the nutritional value of foods directly from food images, but they do not consider how nutritional value changes after cooking. Consequently, VegeNet, a multi-output InceptionV3-based convolutional neural network model has been developed, which estimates the nutritional values of cooked and uncooked vegetables. The explicit use of the cooking state is the main contribution of this work. This deep learning model successfully classifies the food images at 97% accuracy and estimates the nutritional values at 15.30% mean relative error, making it suitable as a visual-based added food assessment solution. This can help users save time and avoid under-reporting problems.

Abstract Accurate calculation of specific spectral properties for NMR is an important step for molecular structure elucidation. Here we report the development of a novel machine learning technique for accurately predicting chemical shifts of both $${^1\mathrm{H}}$$ 1H and $${^{13}\mathrm{C}}$$ 13C nuclei which exceeds DFT-accessible accuracy for $${^{13}\mathrm{C}}$$ 13C and $${^1\mathrm{H}}$$ 1H for a subset of nuclei, while being orders of magnitude more performant. Our method produces estimates of uncertainty, allowing for robust and confident predictions, and suggests future avenues for improved performance.

We demonstrate that particle swarm optimisation (PSO) can be used to solve a variety of problems arising during operation of a digital inspection microscope. This is a use case for the feasibility of heuristics in a real-world product. We show solutions to four measurement problems, all based on PSO. This allows for a compact software implementation solving different problems. We have found that PSO can solve a variety of problems with small software footprints and good results in a real-world embedded system. Notably, in the microscope application, this eliminates the need to return the device to the factory for calibration.

The COVID-19 pandemic caused a shift in teaching practice towards blended learning for many higher education institutions. This led to the rapid adoption of certain digital technologies within existing teaching structures as a means to meet student access needs. This paper is an attempt to summarise and extend pre-COVID-19 pedagogical research to leverage digital immersive technologies for blended teaching in the post-pandemic era. This paper forms both a review of these methodologies and a case study of the I-Ulysses Virtual Learning Environment as an example of a platform that leverages such immersive digital technologies and employs instrumental use of VR. To further clarify, the purpose of the paper is to describe and propose a distance learning solution with immersive VR qualities; this is what the I-Ulysses environment represents, as the main obstacle to learners of site-specific information during the pandemic has been lack of on-site accessibility. Furthermore, this is of key importance, because Joyce’s novel takes place in historical Dublin, where access to the physical location of the story is indispensable to a reader.

Abstract Background Since 2009, IPF patients across Europe are recruited into the eurIPFreg, providing epidemiological data and biomaterials for translational research. Methods The registry data are based on patient and physician baseline and follow-up questionnaires, comprising 1700 parameters. The mid- to long-term objectives of the registry are to provide clues for a better understanding of IPF phenotype sub-clusters, triggering factors and aggravating conditions, regional and environmental characteristics, and of disease behavior and management. Results This paper describes baseline data of 525 IPF subjects recruited from 11/2009 until 10/2016. IPF patients had a mean age of 68.1 years, and seeked medical advice due to insidious dyspnea (90.1%), fatigue (69.2%), and dry coughing (53.2%). A surgical lung biopsy was performed in 32% in 2009, but in only 8% of the cases in 2016, possibly due to increased numbers of cryobiopsy. At the time of inclusion in the eurIPFreg, FVC was 68.4% ± 22.6% of predicted value, DLco ranged at 42.1% ± 17.8% of predicted value (mean value ± SD). Signs of pulmonary hypertension were found in 16.8%. Steroids, immunosuppressants and N-Acetylcysteine declined since 2009, and were replaced by antifibrotics, under which patients showed improved survival (p = 0.001). Conclusions Our data provide important insights into baseline characteristics, diagnostic and management changes as well as outcome data in European IPF patients over time. Trial registration The eurIPFreg and eurIPFbank are listed in ClinicalTrials.gov(NCT02951416).

Nanomechanical sensors that rely on intrinsic resonance frequencies usually present a tradeoff between sensitivity and bandwidth. In this work, the authors realise an optically driven nanorotor featuring high frequency stability and tunability over a large frequency range.

Abstract Background The Chemistry Development Kit (CDK) is a widely used open source cheminformatics toolkit, providing data structures to represent chemical concepts along with methods to manipulate such structures and perform computations on them. The library implements a wide variety of cheminformatics algorithms ranging from chemical structure canonicalization to molecular descriptor calculations and pharmacophore perception. It is used in drug discovery, metabolomics, and toxicology. Over the last 10 years, the code base has grown significantly, however, resulting in many complex interdependencies among components and poor performance of many algorithms. Results We report improvements to the CDK v2.0 since the v1.2 release series, specifically addressing the increased functional complexity and poor performance. We first summarize the addition of new functionality, such atom typing and molecular formula handling, and improvement to existing functionality that has led to significantly better performance for substructure searching, molecular fingerprints, and rendering of molecules. Second, we outline how the CDK has evolved with respect to quality control and the approaches we have adopted to ensure stability, including a code review mechanism. Conclusions This paper highlights our continued efforts to provide a community driven, open source cheminformatics library, and shows that such collaborative projects can thrive over extended periods of time, resulting in a high-quality and performant library. By taking advantage of community support and contributions, we show that an open source cheminformatics project can act as a peer reviewed publishing platform for scientific computing software. Graphical abstract CDK 2.0 provides new features and improved performance

This article proposes a framework that automatically designs classifiers for the early detection of COVID-19 from chest X-ray images. To do this, our approach repeatedly makes use of a heuristic for optimisation to efficiently find the best combination of the hyperparameters of a convolutional deep learning model. The framework starts with optimising a basic convolutional neural network which represents the starting point for the evolution process. Subsequently, at most two additional convolutional layers are added, at a time, to the previous convolutional structure as a result of a further optimisation phase. Each performed phase maximises the the accuracy of the system, thus requiring training and assessment of the new model, which gets gradually deeper, with relevant COVID-19 chest X-ray images. This iterative process ends when no improvement, in terms of accuracy, is recorded. Hence, the proposed method evolves the most performing network with the minimum number of convolutional layers. In this light, we simultaneously achieve high accuracy while minimising the presence of redundant layers to guarantee a fast but reliable model. Our results show that the proposed implementation of such a framework achieves accuracy up to 99.11%, thus being particularly suitable for the early detection of COVID-19.

The role and importance of the identification of natural products are discussed in the perspective of the study of secondary metabolites. The rapid identification of already reported compounds, or structural dereplication, is recognized as a key element in natural product chemistry. The biological taxonomy of metabolite producing organisms, the knowledge of metabolite molecular structures, and the availability of metabolite spectroscopic signatures are considered as the three pillars of structural dereplication. The role and the construction of databases is illustrated by references to the KNApSAcK, UNPD, CSEARCH, and COCONUT databases, and by the importance of calculated taxonomic and spectroscopic data as substitutes for missing or lost original ones. Two NMR-based tools, the PNMRNP database that derives from UNPD, and KnapsackSearch, a database generator that provides taxonomically focused libraries of compounds, are proposed to the community of natural product chemists. The study of the alkaloids from Urceolina peruviana, a plant from the Andes used in traditional medicine for antibacterial and anticancer actions, has given the opportunity to test different approaches to dereplication, favoring the use of publicly available data sources.

Whether quantum physics is universally valid is an open question with far-reaching implications. Intense research is therefore invested into testing the quantum superposition principle with ever heavier and more complex objects. Here we propose a radically new, experimentally viable route towards studies at the quantum-to-classical borderline by probing the orientational quantum revivals of a nanoscale rigid rotor. The proposed interference experiment testifies a macroscopic superposition of all possible orientations. It requires no diffraction grating, uses only a single levitated particle, and works with moderate motional temperatures under realistic environmental conditions. The first exploitation of quantum rotations of a massive object opens the door to new tests of quantum physics with submicron particles and to quantum gyroscopic torque sensors, holding the potential to improve state-of-the-art devices by many orders of magnitude.