Multi-dichotomous sequencing method MDSM for ordering the importance of variants' properties in multi-criteria decision-making

There are plenty of Multi-Criteria Decision-Making (MCDM) methods that help to choose the most suitable solution assessed by several criteria (e.g. Saaty 1990; Simos 1990; Pamučar et al. 2018). They are applied in cases where several scales of different units describe the variants or the variants' properties are represented by linguistic, non-numbered terms. The inherent part of the MCDM algorithms is calculating the weights of the variants' properties, necessary for ordering the variants. If - in a certain problem - there are several properties to consider, sequencing their importance becomes a problem itself. The innovative method of sequencing is proposed in the article based on dichotomous splitting of the properties' importance. If made several times, it leads to the coherent - internally and with the decision-maker's intention - order of the properties' importance. Then the weights of the properties can be calculated with the use of different MCDM methods. The description of the method can be shortened as follows:•Divide the full set of features into two dichotomous subsets of lower and higher importance•Continue dichotomous divisions until there are only the subsets containing one element or subsets containing elements of equal importance.

A quantum synthetic aperture radar image denoising algorithm based on grayscale morphology

The quantum denoising technology efficiently removes noise from images; however, the existing algorithms are only effective for additive noise and cannot remove multiplicative noise, such as speckle noise in synthetic aperture radar (SAR) images. In this paper, based on the grayscale morphology method, a quantum SAR image denoising algorithm is proposed, which performs morphological operations on all pixels simultaneously to remove the noise in the SAR image. In addition, we design a feasible quantum adder to perform cyclic shift operations. Then, quantum circuits for dilation and erosion are designed, and the complete quantum circuit is then constructed. For a 2 n × 2 n quantum SAR image with q grayscale levels, the complexity of our algorithm is O ( n + q ) . Compared with classical algorithms, it achieves exponential improvement and also has polynomial-level improvements than existing quantum algorithms. Finally, the feasibility of our algorithm is validated on IBM Q.

The carbon footprint of predicting CO2 storage capacity in metal-organic frameworks within neural networks

While artificial intelligence drives remarkable progress in natural sciences, its broader societal implications are mostly disregarded. In this study, we evaluate environmental impacts of deep learning in materials science through extensive benchmarking. In particular, a set of diverse neural networks is trained for a given supervised learning task to assess greenhouse gas (GHG) emissions during training and inference phases. A chronological perspective showed diminishing returns, manifesting themselves as a 28% decrease in mean absolute error and nearly a 15,000% increase in the carbon footprint of model training in 2016-2022. By means of up-to-date graphics processing units, it is possible to partially offset the immense growth of GHG emissions. Nonetheless, the practice of employing energy-efficient hardware is overlooked by the materials informatics community, as follows from a literature analysis in the field. On the basis of our findings, we encourage researchers to report GHG emissions together with standard performance metrics.

Genome-scale metabolic network models for industrial microorganisms metabolic engineering: Current advances and future prospects

The construction of high-performance microbial cell factories (MCFs) is the centerpiece of biomanufacturing. However, the complex metabolic regulatory network of microorganisms poses great challenges for the efficient design and construction of MCFs. The genome-scale metabolic network models (GSMs) can systematically simulate the metabolic regulation process of microorganisms in silico, providing effective guidance for the rapid design and construction of MCFs. In this review, we summarized the development status of 16 important industrial microbial GSMs, and further outline the technologies or methods that continuously promote high-quality GSMs construction from five aspects: I) Databases and modeling tools facilitate GSMs reconstruction; II) evolving gap-filling technologies; III) constraint-based model reconstruction; IV) advances in algorithms; and V) developed visualization tools. In addition, we also summarized the applications of GSMs in guiding metabolic engineering from four aspects: I) exploring and explaining metabolic features; II) predicting the effects of genetic perturbations on metabolism; III) predicting the optimal phenotype; IV) guiding cell factories construction in practical experiment. Finally, we discussed the development of GSMs, aiming to provide a reference for efficiently reconstructing GSMs and guiding metabolic engineering.

Systematic approaches to machine learning models for predicting pesticide toxicity

Pesticides play an important role in modern agriculture by protecting crops from pests and diseases. However, the negative consequences of pesticides, such as environmental contamination and adverse effects on human and ecological health, underscore the importance of accurate toxicity predictions. To address this issue, artificial intelligence models have emerged as valuable methods for predicting the toxicity of organic compounds. In this review article, we explore the application of machine learning (ML) for pesticide toxicity prediction. This review provides a detailed summary of recent developments, prediction models, and datasets used for pesticide toxicity prediction. In this analysis, we compared the results of several algorithms that predict the harmfulness of various classes of pesticides. Furthermore, this review article identified emerging trends and areas for future direction, showcasing the transformative potential of machine learning in promoting safer pesticide usage and sustainable agriculture.

Performance of anterior segment OCT-based algorithms in the opportunistic screening for primary angle-closure disease

Purpose

This study aimed to investigate the performance of deep learning algorithms in the opportunistic screening for primary angle-closure disease (PACD) using combined anterior segment parameters.

Methods

This was an observational, cross-sectional hospital-based study. Patients with PACD and healthy controls who underwent comprehensive eye examinations, including gonioscopy and anterior segment optical coherence tomography (ASOCT) examinations under both light and dark conditions, were consecutively enrolled from the Department of Ophthalmology at the Beijing Tongren Hospital between November 2020 and June 2022. The anterior chamber, anterior chamber angle, iris, and lens parameters were assessed using ASOCT. To build the prediction models, backward logistic regression was utilized to select the variables to discriminate patients with PACD from normal participants, and the area under the receiver operating characteristic curve was used to evaluate the efficacy of the opportunistic screening.

Results

The data from 199 patients (199 eyes) were included in the final analysis and divided into two groups: PACD (109 eyes) and controls (90 eyes). Angle opening distance at 500 μm, anterior chamber area, and iris curvature measured in the light condition were included in the final prediction models. The area under the receiver operating characteristic curve was 0.968, with a sensitivity of 91.74 % and a specificity of 91.11 %.

Conclusion

ASOCT-based algorithms showed excellent diagnostic performance in the opportunistic screening for PACD. These results provide a promising basis for future research on the development of an angle-closure probability scoring system for PACD screening.

Development and evaluation of a deep learning-based model for simultaneous detection and localization of rib and clavicle fractures in trauma patients' chest radiographs

Purpose

To develop a rib and clavicle fracture detection model for chest radiographs in trauma patients using a deep learning (DL) algorithm.

Materials and methods

We retrospectively collected 56 145 chest X-rays (CXRs) from trauma patients in a trauma center between August 2008 and December 2016. A rib/clavicle fracture detection DL algorithm was trained using this data set with 991 (1.8%) images labeled by experts with fracture site locations. The algorithm was tested on independently collected 300 CXRs in 2017. An external test set was also collected from hospitalized trauma patients in a regional hospital for evaluation. The receiver operating characteristic curve with area under the curve (AUC), accuracy, sensitivity, specificity, precision, and negative predictive value of the model on each test set was evaluated. The prediction probability on the images was visualized as heatmaps.

Results

The trained DL model achieved an AUC of 0.912 (95% CI 87.8 to 94.7) on the independent test set. The accuracy, sensitivity, and specificity on the given cut-off value are 83.7, 86.8, and 80.4, respectively. On the external test set, the model had a sensitivity of 88.0 and an accuracy of 72.5. While the model exhibited a slight decrease in accuracy on the external test set, it maintained its sensitivity in detecting fractures.

Conclusion

The algorithm detects rib and clavicle fractures concomitantly in the CXR of trauma patients with high accuracy in locating lesions through heatmap visualization.

Phylogenetic network classes through the lens of expanding covers

It was recently shown that a large class of phylogenetic networks, the 'labellable' networks, is in bijection with the set of 'expanding' covers of finite sets. In this paper, we show how several prominent classes of phylogenetic networks can be characterised purely in terms of properties of their associated covers. These classes include the tree-based, tree-child, orchard, tree-sibling, and normal networks. In the opposite direction, we give an example of how a restriction on the set of expanding covers can define a new class of networks, which we call 'spinal' phylogenetic networks.

[Artificial intelligence and acute kidney injury]

Digitalization is increasingly finding its way into intensive care units and with it artificial intelligence (AI) for critically ill patients. One promising area for the use of AI is in the field of acute kidney injury (AKI). The use of AI is primarily focused on the prediction of AKI, but further approaches are also being used to classify existing AKI into different phenotypes. Different AI models are used for prediction. The area under the receiver operating characteristic curve values (AUROC) achieved with these models vary and are influenced by several factors, such as the prediction time and the definition of AKI. Most models have an AUROC between 0.650 and 0.900, with lower values for predictions further into the future and when applying Acute Kidney Injury Network (AKIN) instead of KDIGO criteria. Classification into phenotypes already makes it possible to categorize patients into groups with different risks of mortality or requirement of renal replacement therapy (RRT), but the etiologies or therapeutic consequences derived from this are still lacking. However, all the models suffer from AI-specific shortcomings. The use of large databases does not make it possible to promptly include recent changes in therapy and the implementation of new biomarkers in a relevant proportion. For this reason, serum creatinine and urinary output, with their known limitations, dominate current AI models for prediction impairing the performance of the current models. On the other hand, the increasingly complex models no longer allow physicians to understand the basis on which the warning of a threatening AKI is calculated and subsequent initiation of therapy should take place. The successful use of AIs in routine clinical practice will be highly determined by the trust of the physicians in the systems and overcoming the aforementioned weaknesses. However, the clinician will remain irreplaceable as the decisive authority for critically ill patients by combining measurable and nonmeasurable parameters.

[Nerve injuries due to fractures in childhood : Primarily and secondarily on the upper extremity]

The approach for nerve injuries in children in the context of fractures of the upper extremities is inconsistent in the literature. The underlying mostly retrospective studies do not usually consider the potential diagnostics. The frequency of nerve injuries with a clear need for reconstructive surgery is sometimes estimated so differently that precedent-setting errors in these studies must be assumed; however, as 10-20% of pediatric fractures near the elbow show primary or secondary nerve lesions, timely and appropriate further treatment is necessary. An overview concerning diagnostic tools with an explanation of potential results and an algorithm with a timeline for diagnostic and therapeutic management are presented. Good results after nerve lesions can only be achieved when timely diagnostics without delay and correct detection of axonal lesions which benefit from surgical treatment are carried out.

AI-powered COVID-19 forecasting: a comprehensive comparison of advanced deep learning methods

Objectives

The coronavirus disease 2019 (COVID-19) pandemic continues to pose significant challenges to the public health sector, including that of the United Arab Emirates (UAE). The objective of this study was to assess the efficiency and accuracy of various deep-learning models in forecasting COVID-19 cases within the UAE, thereby aiding the nation's public health authorities in informed decision-making.

Methods

This study utilized a comprehensive dataset encompassing confirmed COVID-19 cases, demographic statistics, and socioeconomic indicators. Several advanced deep learning models, including long short-term memory (LSTM), bidirectional LSTM, convolutional neural network (CNN), CNN-LSTM, multilayer perceptron, and recurrent neural network (RNN) models, were trained and evaluated. Bayesian optimization was also implemented to fine-tune these models.

Results

The evaluation framework revealed that each model exhibited different levels of predictive accuracy and precision. Specifically, the RNN model outperformed the other architectures even without optimization. Comprehensive predictive and perspective analytics were conducted to scrutinize the COVID-19 dataset.

Conclusion

This study transcends academic boundaries by offering critical insights that enable public health authorities in the UAE to deploy targeted data-driven interventions. The RNN model, which was identified as the most reliable and accurate for this specific context, can significantly influence public health decisions. Moreover, the broader implications of this research validate the capability of deep learning techniques in handling complex datasets, thus offering the transformative potential for predictive accuracy in the public health and healthcare sectors.

A systematic review and meta-analysis

BACKGROUND

There is a relative lack of research evaluating the outcomes when treatment guidelines or algorithms for psychotic disorders are followed. This systematic review and meta-analysis determined the response rates to antipsychotic medications at different stages of these algorithms and whether these response rates differ in first episode cohorts.

METHODS

Data sources: A systematic search strategy was conducted across four databases PubMed, EMBASE, PsycINFO (Ovid) and CINAHL. Studies that had sequential trials of different antipsychotic medications were included. A meta-analysis of proportions was performed using random effects models and sub-group analysis in first episode psychosis studies.

RESULTS

Of the 4078 unique articles screened, fourteen articles, from nine unique studies, were eligible and included 2522 participants. The proportion who experienced a response to any antipsychotic in the first stage of an algorithm was 0.53 (95 % C.I.:0.38,0.68) and this decreased to 0.26 (95 % C.I.:0.15,0.39) in the second stage. When clozapine was used in the third stage, the proportion that achieved a response was 0.43 (95 % C.I. 0.19, 0.69) compared to 0.26 (95 % C.I.:0.05,0.54) if a different antipsychotic was used. Four studies included 907 participants with a first episode of psychosis and the proportions that achieved a response were: 1st stage: 0.63 (95 % C.I.: 0.45, 0.79); 2nd stage: 0.34 (95 % C.I.:0.16,0.55); clozapine 3rd stage: 0.45 (95 % C.I.:0.0,0.97), different antipsychotic 3rd stage: 0.15 (95 % C.I.,0.01,0.37).

DISCUSSION

These findings support the recommendation to have a trial of clozapine after two other antipsychotic medications have been found to be ineffective.

Relevance of sleep for wellness: New trends in using artificial intelligence and machine learning

Sleep and well-being have been intricately linked, and sleep hygiene is paramount for developing mental well-being and resilience. Although widespread, sleep disorders require elaborate polysomnography laboratory and patient-stay with sleep in unfamiliar environments. Current technologies have allowed various devices to diagnose sleep disorders at home. However, these devices are in various validation stages, with many already receiving approvals from competent authorities. This has captured vast patient-related physiologic data for advanced analytics using artificial intelligence through machine and deep learning applications. This is expected to be integrated with patients' Electronic Health Records and provide individualized prescriptive therapy for sleep disorders in the future.

Assessing robustness and generalization of a deep neural network for brain MS lesion segmentation on real-world data

OBJECTIVES

Evaluate the performance of a deep learning (DL)-based model for multiple sclerosis (MS) lesion segmentation and compare it to other DL and non-DL algorithms.

METHODS

This ambispective, multicenter study assessed the performance of a DL-based model for MS lesion segmentation and compared it to alternative DL- and non-DL-based methods. Models were tested on internal (n = 20) and external (n = 18) datasets from Latin America, and on an external dataset from Europe (n = 49). We also examined robustness by rescanning six patients (n = 6) from our MS clinical cohort. Moreover, we studied inter-human annotator agreement and discussed our findings in light of these results. Performance and robustness were assessed using intraclass correlation coefficient (ICC), Dice coefficient (DC), and coefficient of variation (CV).

RESULTS

Inter-human ICC ranged from 0.89 to 0.95, while spatial agreement among annotators showed a median DC of 0.63. Using expert manual segmentations as ground truth, our DL model achieved a median DC of 0.73 on the internal, 0.66 on the external, and 0.70 on the challenge datasets. The performance of our DL model exceeded that of the alternative algorithms on all datasets. In the robustness experiment, our DL model also achieved higher DC (ranging from 0.82 to 0.90) and lower CV (ranging from 0.7 to 7.9%) when compared to the alternative methods.

CONCLUSION

Our DL-based model outperformed alternative methods for brain MS lesion segmentation. The model also proved to generalize well on unseen data and has a robust performance and low processing times both on real-world and challenge-based data.

CLINICAL RELEVANCE STATEMENT

Our DL-based model demonstrated superior performance in accurately segmenting brain MS lesions compared to alternative methods, indicating its potential for clinical application with improved accuracy, robustness, and efficiency.

KEY POINTS

• Automated lesion load quantification in MS patients is valuable; however, more accurate methods are still necessary. • A novel deep learning model outperformed alternative MS lesion segmentation methods on multisite datasets. • Deep learning models are particularly suitable for MS lesion segmentation in clinical scenarios.

Clustering the cortical laminae: in vivo parcellation

The laminar microstructure of the cerebral cortex has distinct anatomical characteristics of the development, function, connectivity, and even various pathologies of the brain. In recent years, multiple neuroimaging studies have utilized magnetic resonance imaging (MRI) relaxometry to visualize and explore this intricate microstructure, successfully delineating the cortical laminar components. Despite this progress, T1 is still primarily considered a direct measure of myeloarchitecture (myelin content), rather than a probe of tissue cytoarchitecture (cellular composition). This study aims to offer a robust, whole-brain validation of T1 imaging as a practical and effective tool for exploring the laminar composition of the cortex. To do so, we cluster complex microstructural cortical datasets of both human (N = 30) and macaque (N = 1) brains using an adaptation of an algorithm for clustering cell omics profiles. The resulting cluster patterns are then compared to established atlases of cytoarchitectonic features, exhibiting significant correspondence in both species. Lastly, we demonstrate the expanded applicability of T1 imaging by exploring some of the cytoarchitectonic features behind various unique skillsets, such as musicality and athleticism.

The proton arc therapy treatment planning problem is NP-Hard

Proton arc therapy (PAT) is an advanced radiotherapy technique using charged particles in which the radiation device rotates continuously around the patient while irradiating the tumor. Compared to conventional, fixed-angle beam delivery mode, proton arc therapy has the potential to further improve the quality of cancer treatment by delivering accurate radiation dose to tumors while minimizing damage to surrounding healthy tissues. However, the computational complexity of treatment planning in PAT raises challenges as to its effective implementation. In this paper, we demonstrate that designing a PAT plan through algorithmic methods is a NP-hard problem (in fact, NP-complete), where the problem size is determined by the number of discrete irradiation angles from which the radiation can be delivered. This finding highlights the inherent complexity of PAT treatment planning and emphasizes the need for efficient algorithms and heuristics to address the challenges associated with optimizing the delivery of radiation doses in this context.

Whole Spine Segmentation Using Object Detection and Semantic Segmentation

OBJECTIVE

Virtual and augmented reality have enjoyed increased attention in spine surgery. Preoperative planning, pedicle screw placement, and surgical training are among the most studied use cases. Identifying osseous structures is a key aspect of navigating a 3-dimensional virtual reconstruction. To automate the otherwise time-consuming process of labeling vertebrae on each slice individually, we propose a fully automated pipeline that automates segmentation on computed tomography (CT) and which can form the basis for further virtual or augmented reality application and radiomic analysis.

METHODS

Based on a large public dataset of annotated vertebral CT scans, we first trained a YOLOv8m (You-Only-Look-Once algorithm, Version 8 and size medium) to detect each vertebra individually. On the then cropped images, a 2D-U-Net was developed and externally validated on 2 different public datasets.

RESULTS

Two hundred fourteen CT scans (cervical, thoracic, or lumbar spine) were used for model training, and 40 scans were used for external validation. Vertebra recognition achieved a mAP50 (mean average precision with Jaccard threshold of 0.5) of over 0.84, and the segmentation algorithm attained a mean Dice score of 0.75 ± 0.14 at internal, 0.77 ± 0.12 and 0.82 ± 0.14 at external validation, respectively.

CONCLUSION

We propose a 2-stage approach consisting of single vertebra labeling by an object detection algorithm followed by semantic segmentation. In our externally validated pilot study, we demonstrate robust performance for our object detection network in identifying individual vertebrae, as well as for our segmentation model in precisely delineating the bony structures.

[Treatment strategy for an unstable chest wall after cardiopulmonary resuscitation]

Every year ca. 60,000 people in Germany undergo cardiopulmonary resuscitation (CPR). The two most frequent underlying causes are of cardiopulmonary and traumatic origin. According to the current CPR guidelines chest compressions should be performed in the middle of the sternum with a pressure frequency of 100-120/min and to a depth of 5-6 cm. In contrast to trauma patients where different injury patterns can arise depending on the accident mechanism, both the type of trauma and the injury pattern are similar in patients after CPR due to repetitive thorax compression. It is known that an early reconstruction of the thoracic wall and the restoration of the physiological breathing mechanics in trauma patients with unstable thoracic injuries reduce the rates of pneumonia and weaning failure and shorten the length of stay in the intensive care unit. As a result, it is increasingly being propagated that an unstable thoracic injury as a result of CPR should also be subjected to surgical treatment as soon as possible. In the hospital of the authors an algorithm was formulated based on clinical experience and the underlying evidence in a traumatological context and a surgical treatment strategy was designed, which is presented and discussed taking the available evidence into account.

Improving the TIR3B oncological stratification: try to bridge the gap through a comprehensive presurgical algorithm

PURPOSE

Indeterminate cytology still puzzles clinicians, due to its wide range of oncological risks. According to the Italian SIAPEC-IAP classification, TIR3B cytology holds up to 30% of thyroid cancer, which justifies the surgical indication, even if more than half of cases do not result in a positive histology. The study aim is to identify potential clinical, ultrasound or cytological features able to improve the surgical indication.

METHODS

Retrospective analysis. A consecutive series of TIR3B nodules referred to the Endocrine Unit of Careggi Hospital from 1st May 2014 to 31st December 2021 was considered for the exploratory analysis (Phase 1). Thereafter, a smaller confirmatory sample of consecutive TIR3B diagnosed and referred to surgery from 1st January 2022 to 31st June 2022 was considered to verify the algorithm (Phase 2). The main clinical, ultrasound and cytological features have been collected. A comprehensive stepwise logistic regression was applied to build a prediction algorithm. The histological results represented the final outcome.

RESULTS

Of 599 TIR3B nodules referred to surgery, 451 cases were included in the exploratory analysis. A final score > 14.5 corresponded to an OR = 4.98 (95% CI 3.24-7.65, p < 0.0001) and showed a PPV and NPV of 57% and 79%, respectively. The Phase 2 analysis on a confirmatory sample of 58 TIR3B cytology confirmed that a threshold of 14.5 points has a comparable PPV and NPV of 53% and 80%, respectively.

CONCLUSIONS

A predictive algorithm which considers the main clinical, US and cytological features can significantly improve the oncological stratification of TIR3B cytology.

Fully automated computational measurement of noise in positron emission tomography

OBJECTIVES

To introduce an automated computational algorithm that estimates the global noise level across the whole imaging volume of PET datasets.

METHODS

[18F]FDG PET images of 38 patients were reconstructed with simulated decreasing acquisition times (15-120 s) resulting in increasing noise levels, and with block sequential regularized expectation maximization with beta values of 450 and 600 (Q.Clear 450 and 600). One reader performed manual volume-of-interest (VOI) based noise measurements in liver and lung parenchyma and two readers graded subjective image quality as sufficient or insufficient. An automated computational noise measurement algorithm was developed and deployed on the whole imaging volume of each reconstruction, delivering a single value representing the global image noise (Global Noise Index, GNI). Manual noise measurement values and subjective image quality gradings were compared with the GNI.

RESULTS

Irrespective of the absolute noise values, there was no significant difference between the GNI and manual liver measurements in terms of the distribution of noise values (p = 0.84 for Q.Clear 450, and p = 0.51 for Q.Clear 600). The GNI showed a fair to moderately strong correlation with manual noise measurements in liver parenchyma (r = 0.6 in Q.Clear 450, r = 0.54 in Q.Clear 600, all p < 0.001), and a fair correlation with manual noise measurements in lung parenchyma (r = 0.52 in Q.Clear 450, r = 0.33 in Q.Clear 600, all p < 0.001). Classification performance of the GNI for subjective image quality was AUC 0.898 for Q.Clear 450 and 0.919 for Q.Clear 600.

CONCLUSION

An algorithm provides an accurate and meaningful estimation of the global noise level encountered in clinical PET imaging datasets.

CLINICAL RELEVANCE STATEMENT

An automated computational approach that measures the global noise level of PET imaging datasets may facilitate quality standardization and benchmarking of clinical PET imaging within and across institutions.

KEY POINTS

• Noise is an important quantitative marker that strongly impacts image quality of PET images. • An automated computational noise measurement algorithm provides an accurate and meaningful estimation of the global noise level encountered in clinical PET imaging datasets. • An automated computational approach that measures the global noise level of PET imaging datasets may facilitate quality standardization and benchmarking as well as protocol harmonization.

Development and applications of genome-scale metabolic network models

The genome-scale metabolic network model is an effective tool for characterizing the gene-protein-response relationship in the entire metabolic pathway of an organism. By combining various algorithms, the genome-scale metabolic network model can effectively simulate the influence of a specific environment on the physiological state of cells, optimize the culture conditions of strains, and predict the targets of genetic modification to achieve targeted modification of strains. In this review, we summarize the whole process of model building, sort out the various tools that may be involved in the model building process, and explain the role of various algorithms in model analysis. In addition, we also summarized the application of GSMM in network characteristics, cell phenotypes, metabolic engineering, etc. Finally, we discuss the current challenges facing GSMM.

Exploring the transformative power of AI in art through a circular economy lens. A systematic literature review

Today, technology and sustainability are two strategic axes for the development of any industry. Art is no exception and embodies both principles. Artificial intelligence (AI) is driving the art world forwards with its applications and algorithms. Additionally, the circular economy (CE) is concerned with resources and the environment in this context. The objective of the present work is to provide an overview of the current state of research on the application of AI in the art world and an analysis of how CE principles are being incorporated, considering the interactions between AI and the CE. To this end, a systematic review of the literature is carried out in which 60 articles related to the subject are selected, analysed, and classified, highlighting the lines of research addressed. The assessment of the current state of research on the subject concludes with the four main axes of classification of works. The first line is related to AI generative content in art, addressing issues of content creation, image and painting, video, and theatre. The second line is related to AI applications for art industry production, considering the sustainability of the supply chain. The third line focuses on how the CE is being applied to art, while the fourth line focuses on other relevant aspects analysed, such as training and design. The topic is still incipient, mandating further research to study the full potential of AI and the CE in the world of art.

Comparison of different methods of screening for preterm pre-eclampsia: cohort study

OBJECTIVE

To compare the predictive performance for pre-eclampsia (PE) of three different first-trimester mathematical models of screening, which combine maternal risk factors with mean arterial pressure (MAP), uterine artery pulsatility index (UtA-PI) and serum placental growth factor (PlGF), and two risk scoring systems, based on NICE and ACOG recommendations.

METHODS

This was a prospective cohort study performed in eight fetal-medicine units in five different regions of Spain between September 2017 and December 2019. All pregnant women with singleton pregnancies and non-malformed live fetuses attending their routine ultrasound examination at 11+0 to 13+6 weeks' gestation were invited to participate in the study. Maternal characteristics and medical history were recorded and measurements of MAP, UtA-PI, serum PlGF and pregnancy associated plasma protein-A (PAPP-A) were converted into multiples of the median (MoM). Risks for term, preterm-PE (< 37 weeks' gestation) and early-PE (< 34 weeks' gestation) were calculated according to the FMF competing risks model, the Crovetto et al., logistic regression model, and Serra et al., Gaussian model. Patient classification based on NICE and ACOG guidelines was also performed. We estimated detection rates (DR) with their 95% confidence intervals (CIs) at a fixed 10% screen positive rate (SPR), as well as the area under the receiver operating characteristic curve (AUROC) for preterm-PE, early-PE, and all PE for the three mathematical models. For the scoring systems, we calculated DR and SPR. Risk calibration was also assessed.

RESULTS

The study population comprised of 10,110 singleton pregnancies, including 32 (0.3%) that developed early-PE, 72 (0.7%) that developed preterm-PE and 230 (2.3%) of any PE. At fixed 10% SPR, the FMF, Crovetto et al., and Serra et al., detected 82.7% (95% CI, 69.6 to 95.8%), 73.8% (95% CI, 58.7 to 88.9%), and 79.8% (95% CI, 66.1 to 93.5%) of early-PE; 72.7% (95% CI, 62.9 to 82.6%), 69.2% (95% CI, 58.8 to 79.6%), and 74.1% (95% CI, 64.2 to 83.9%) of preterm-PE and 55.1% (95% CI, 48.8 to 61.4%), 47.1% (95% CI, 40.6 to 53.5%), and 53.9% (95% CI, 47.4 to 60.4%) of all PE, respectively. The best correlation between predicted and observed cases was achieved by the FMF model, with an AUROC of 0.911 (95% CI, 0.879 to 0.943), a slope of 0.983 (95% CI, 0.846-1.120) and an intercept of 0.154 (95% CI, -0.091 to 0.397). The NICE criteria identified 46.7% (95% CI, 35.3 to 58.0%) of preterm-PE at 11% SPR and ACOG criteria identified 65.9% (95% CI, 55.4 to 76.4%) of preterm-PE at 33.8% SPR.

CONCLUSIONS

The best performance of screening for preterm-PE is achieved by mathematical models that combine maternal factors with MAP, UtA-PI and PlGF, as compared to risk-scoring systems like NICE or ACOG criteria. While all three algorithms show similar results in terms of overall prediction, the FMF model showed the best performance at the individual level. This article is protected by copyright. All rights reserved.

A study on a recommendation algorithm based on spectral clustering and GRU

With the development of e-commerce, the importance of recommendation algorithms has significantly increased. However, traditional recommendation systems struggle to address issues such as data sparsity and cold start. This article proposes an optimization method for a recommendation system based on spectral clustering (SC) and gated recurrent unit (GRU), named the GRU-KSC algorithm. Firstly, this paper improves the original spectral clustering algorithm by introducing Kmc2, proposing a novel spectral clustering recommendation algorithm (K-means++ SC, KSC) based on the existing SC algorithm. Secondly, building upon the original GRU model, the paper presents a hybrid recommendation algorithm (Hybrid GRU, HGRU) capable of capturing long-term user interests for a more personalized recommendation. Experiments conducted on real datasets demonstrate that our method outperforms existing benchmark methods in terms of accuracy and robustness.

Concordance and generalization of an AI algorithm with real-world clinical data in the pre-omicron and omicron era

All viruses, including SARS-CoV-2, the virus responsible for COVID-19, continue to evolve, which can lead to new variants. The objective of this study is to assess the agreement between real-world clinical data and an algorithm that utilizes laboratory markers and age to predict the progression of disease severity in COVID-19 patients during the pre-Omicron and Omicron variant periods. The study evaluated the performance of a deep learning (DL) algorithm in predicting disease severity scores for COVID-19 patients using data from the USA, Spain, and Turkey (Ankara City Hospital (ACH) data set). The algorithm was developed and validated using pre-Omicron era data and was tested on both pre-Omicron and Omicron-era data. The predictions were compared to the actual clinical outcomes using a multidisciplinary approach. The concordance index values for all datasets ranged from 0.71 to 0.81. In the ACH cohort, a negative predictive value (NPV) of 0.78 or higher was observed for severe patients in both the pre-Omicron and Omicron eras, which is consistent with the algorithm's performance in the development cohort.

Advancements in Raman imaging for nanoplastic analysis: Challenges, algorithms and future Perspectives

BACKGROUND

While the concept of microplastic (<5 mm) is well-established, emergence of nanoplastics (<1000 nm) as a new contaminant presents a recent and evolving challenge. The field of nanoplastic research remains in its early stages, and its progress is contingent upon the development of reliable and practical analytical methods, which are currently lacking. This review aims to address the intricacies of nanoplastic analysis by providing a comprehensive overview on the application of advanced imaging techniques, with a particular focus on Raman imaging, for nanoplastic identification and simultaneous visualisation towards quantification.

RESULTS

Although Raman imaging via hyper spectrum is a potentially powerful tool to analyse nanoplastics, several challenges should be overcome. The first challenge lies in the weak Raman signal of nanoplastics. To address this, effective sample preparation and signal enhancement techniques can be implemented, such as by analysing the hyper spectrum that contains hundred-to-thousand spectra, rather than a single spectrum. Second challenge is the complexity of Raman hyperspectral matrix with dataset size at megabyte (MB) or even bigger, which can be adopted using different algorithms ranging from image merging to multivariate analysis of chemometrics. Third challenge is the laser size that hinders the visualisation of small nanoplastics due to the laser diffraction (λ/2NA, ∼300 nm), which can be solved with involving the use of super-resolution. Signal processing, such as colour off-setting, Gaussian fitting (via deconvolution), and re-focus or image re-construction, are reviewed herein, which show a great promise for breaking through the diffraction limit.

SIGNIFICANCE

Overall, current studies along with further validation are imperative to refine these approaches and enhance the reliability, not only for nanoplastics research but also for broader investigations in the realm of nanomaterials.

A foundation for evaluating the surgical artificial intelligence literature

With increasing growth in applications of artificial intelligence (AI) in surgery, it has become essential for surgeons to gain a foundation of knowledge to critically appraise the scientific literature, commercial claims regarding products, and regulatory and legal frameworks that govern the development and use of AI. This guide offers surgeons a framework with which to evaluate manuscripts that incorporate the use of AI. It provides a glossary of common terms, an overview of prerequisite knowledge to maximize understanding of methodology, and recommendations on how to carefully consider each element of a manuscript to assess the quality of the data on which an algorithm was trained, the appropriateness of the methodological approach, the potential for reproducibility of the experiment, and the applicability to surgical practice, including considerations on generalizability and scalability.

The accuracy of fully-automated algorithms for the surveillance of central venous catheter-related bloodstream infection in hospitalised patients

BACKGROUND

Continuous surveillance for healthcare-associated infections such as central venous catheter-related bloodstream infections (CVC-BSI) is crucial for prevention. However, traditional surveillance methods are resource-intensive and prone to bias. This study aimed to develop and validate fully-automated surveillance algorithms for CVC-BSI.

METHODS

Two algorithms were developed using electronic health record data from 1000 admissions with a positive blood culture (BCx) at Karolinska University Hospital from 2017: (1) Combining microbiological findings in BCx and CVC cultures with BSI symptoms; (2) Only using microbiological findings. These algorithms were validated in 5170 potential CVC-BSI-episodes from all admissions in 2018-2019, and results extrapolated to all potential CVC-BSI-episodes within this period (n = 181,354). The reference standard was manual record review according to ECDC's definition of microbiologically confirmed CVC-BSI (CRI3-CVC).

RESULTS

In the potential CVC-BSI-episodes, 51 fulfilled ECDC's definition and the algorithms identified 47 and 49 episodes as CVC-BSI, respectively. Both algorithms performed well in assessing CVC-BSI. Overall, algorithm 2 performed slightly better with in the total period a sensitivity of 0.880 (95%-CI 0.783-0.959), specificity of 1.000 (95%-CI 0.999-1.000), PPV of 0.918 (95%-CI 0.833-0.981) and NPV of 1.000 (95%-CI 0.999-1.000). Incidence according to the reference and algorithm 2 was 0.33 and 0.31 per 1000 in-patient hospital-days, respectively.

CONCLUSIONS

Both fully-automated surveillance algorithms for CVC-BSI performed well and could effectively replace manual surveillance. The simpler algorithm, using only microbiology data, is suitable when BCx testing adheres to recommendations, otherwise the algorithm using symptom data might be required. Further validation in other settings is necessary to assess the algorithms' generalisability.

The validity of electronic health data for measuring smoking status: a systematic review and meta-analysis

BACKGROUND

Smoking is a risk factor for many chronic diseases. Multiple smoking status ascertainment algorithms have been developed for population-based electronic health databases such as administrative databases and electronic medical records (EMRs). Evidence syntheses of algorithm validation studies have often focused on chronic diseases rather than risk factors. We conducted a systematic review and meta-analysis of smoking status ascertainment algorithms to describe the characteristics and validity of these algorithms.

METHODS

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were followed. We searched articles published from 1990 to 2022 in EMBASE, MEDLINE, Scopus, and Web of Science with key terms such as validity, administrative data, electronic health records, smoking, and tobacco use. The extracted information, including article characteristics, algorithm characteristics, and validity measures, was descriptively analyzed. Sources of heterogeneity in validity measures were estimated using a meta-regression model. Risk of bias (ROB) in the reviewed articles was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 tool.

RESULTS

The initial search yielded 2086 articles; 57 were selected for review and 116 algorithms were identified. Almost three-quarters (71.6%) of algorithms were based on EMR data. The algorithms were primarily constructed using diagnosis codes for smoking-related conditions, although prescription medication codes for smoking treatments were also adopted. About half of the algorithms were developed using machine-learning models. The pooled estimates of positive predictive value, sensitivity, and specificity were 0.843, 0.672, and 0.918 respectively. Algorithm sensitivity and specificity were highly variable and ranged from 3 to 100% and 36 to 100%, respectively. Model-based algorithms had significantly greater sensitivity (p = 0.006) than rule-based algorithms. Algorithms for EMR data had higher sensitivity than algorithms for administrative data (p = 0.001). The ROB was low in most of the articles (76.3%) that underwent the assessment.

CONCLUSIONS

Multiple algorithms using different data sources and methods have been proposed to ascertain smoking status in electronic health data. Many algorithms had low sensitivity and positive predictive value, but the data source influenced their validity. Algorithms based on machine-learning models for multiple linked data sources have improved validity.

Exploring the potential of ChatGPT in the peer review process: An observational study

BACKGROUND

Peer review is the established method for evaluating the quality and validity of research manuscripts in scholarly publishing. However, scientific peer review faces challenges as the volume of submitted research has steadily increased in recent years. Time constraints and peer review quality assurance can place burdens on reviewers, potentially discouraging their participation. Some artificial intelligence (AI) tools might assist in relieving these pressures. This study explores the efficiency and effectiveness of one of the artificial intelligence (AI) chatbots, ChatGPT (Generative Pre-trained Transformer), in the peer review process.

METHODS

Twenty-one peer-reviewed research articles were anonymised to ensure unbiased evaluation. Each article was reviewed by two humans and by versions 3.5 and 4.0 of ChatGPT. The AI was instructed to provide three positive and three negative comments on the articles and recommend whether they should be accepted or rejected. The human and AI results were compared using a 5-point Likert scale to determine the level of agreement. The correlation between ChatGPT responses and the acceptance or rejection of the papers was also examined.

RESULTS

Subjective review similarity between human reviewers and ChatGPT showed a mean score of 3.6/5 for ChatGPT 3.5 and 3.76/5 for ChatGPT 4.0. The correlation between human and AI review scores was statistically significant for ChatGPT 3.5, but not for ChatGPT 4.0.

CONCLUSION

ChatGPT can complement human scientific peer review, enhancing efficiency and promptness in the editorial process. However, a fully automated AI review process is currently not advisable, and ChatGPT's role should be regarded as highly constrained for the present and near future.

FastPtx: a versatile toolbox for rapid, joint design of pTx RF and gradient pulses using Pytorch's autodifferentiation

OBJECTIVE

With modern optimization methods, free optimization of parallel transmit pulses together with their gradient waveforms can be performed on-line within a short time. A toolbox which uses PyTorch's autodifferentiation for simultaneous optimization of RF and gradient waveforms is presented and its performance is evaluated.

METHODS

MR measurements were performed on a 9.4T MRI scanner using a 3D saturated single-shot turboFlash sequence for [Formula: see text] mapping. RF pulse simulation and optimization were done using a Python toolbox and a dedicated server. An RF- and Gradient pulse design toolbox was developed, including a cost function to balance different metrics and respect hardware and regulatory limits. Pulse performance was evaluated in GRE and MPRAGE imaging. Pulses for non-selective and for slab-selective excitation were designed.

RESULTS

Universal pulses for non-selective excitation reduced the flip angle error to an NRMSE of (12.3±1.7)% relative to the targeted flip angle in simulations, compared to (42.0±1.4)% in CP mode. The tailored pulses performed best, resulting in a narrow flip angle distribution with NRMSE of (8.2±1.0)%. The tailored pulses could be created in only 66 s, making it feasible to design them during an experiment. A 90° pulse was designed as preparation pulse for a satTFL sequence and achieved a NRMSE of 7.1%. We showed that both MPRAGE and GRE imaging benefited from the pTx pulses created with our toolbox.

CONCLUSION

The pTx pulse design toolbox can freely optimize gradient and pTx RF waveforms in a short time. This allows for tailoring high-quality pulses in just over a minute.

SWOffinder: Efficient and versatile search of CRISPR off-targets with bulges by Smith-Waterman alignment

CRISPR/Cas9 technology is revolutionizing the field of gene editing. While this technology enables the targeting of any gene, it may also target unplanned loci, termed off-target sites (OTS), which are a few mismatches, insertions, and deletions from the target. While existing methods for finding OTS up to a given mismatch threshold are efficient, other methods considering insertions and deletions are limited by long runtimes, incomplete OTS lists, and partial support of versatile thresholds. Here, we developed SWOffinder, an efficient method based on Smith-Waterman alignment to find all OTS up to some edit distance. We implemented an original trace-back approach to find OTS under versatile criteria, such as separate limits on the number of insertions, deletions, and mismatches. Compared to state-of-the-art methods, only SWOffinder finds all OTS in the genome in just a few minutes. SWOffinder enables accurate and efficient genomic search of OTS, which will lead to safer gene editing.

A comparative evaluation of nature-inspired algorithms for feature selection problems

Feature selection is a critical component of machine learning and data mining which addresses challenges like irrelevance, noise, redundancy in large-scale data etc., which often result in the curse of dimensionality. This study employs a K-nearest neighbour wrapper to implement feature selection using six nature-inspired algorithms, derived from human behaviour and mammal-inspired techniques. Evaluated on six real-world datasets, the study aims to compare the performance of these algorithms in terms of accuracy, feature count, fitness, convergence and computational cost. The findings underscore the efficacy of the Human Learning Optimization, Poor and Rich Optimization and Grey Wolf Optimizer algorithms across multiple performance metrics. For instance, for mean fitness, Human Learning Optimization outperforms the others, followed by Poor and Rich Optimization and Harmony Search. The study suggests the potential of human-inspired algorithms, particularly Poor and Rich Optimization, in robust feature selection without compromising classification accuracy.

Evaluating consumer self-medication practices, pharmaceutical care services, and pharmacy selection: a quantitative study

BACKGROUND

The primary objectives of this study were the evaluation of consumer self-medication practices, the assessment of pharmaceutical care provided by pharmacy employees, and the analysis of consumer satisfaction with such care. The research was also aimed at examining the main criteria that consumers consider important when selecting a pharmacy in Armenia.

METHODS

The survey was based on an anonymous questionnaire and carried out between March 2020 and November 2021. It was aimed at providing a comprehensive assessment of pharmaceutical care services and consumer pharmacy choice by investigating two distinct groups: pharmacy consumers and pharmacy employees.

RESULTS

The research reveals that many residents in Armenia engage in self-medication without consulting professional sources, which can lead to potential risks and result in dangerous consequences. This is partly due to a lack of trust in pharmacy employees, which is primarily due to their inability to provide adequate information and advice. This study highlights a significant need for improvement in the quality of service provided by pharmacy employees. Despite these challenges, the majority of consumers reported having a preferred pharmacy, and that employee knowledge is the most important criterion when choosing a pharmacy.

CONCLUSIONS

Consumer distrust, in this context, is based on the incomplete knowledge or incompetency of pharmacy employees. Collective actions should be taken to improve the role of pharmacy employees and consequently improve the public trust in them, which can ensure better control of self-medication and reduce the instances of mistreatment.

Selection of reference genes in liproxstatin-1-treated K562 Leukemia cells via RT-qPCR and RNA sequencing

BACKGROUND

Reverse transcription quantitative polymerase chain reaction (RT-qPCR) can accurately detect relative gene expression levels in biological samples. However, widely used reference genes exhibit unstable expression under certain conditions.

METHODS AND RESULTS

Here, we compared the expression stability of eight reference genes (RPLP0, RPS18, RPL13, EEF1A1, β-actin, GAPDH, HPRT1, and TUBB) commonly used in liproxstatin-1 (Lip-1)-treated K562 cells using RNA-sequencing and RT-qPCR. The expression of EEF1A1, ACTB, GAPDH, HPRT1, and TUBB was considerably lower in cells treated with 20 μM Lip-1 than in the control, and GAPDH also showed significant downregulation in the 10 μM Lip-1 group. Meanwhile, when we used geNorm, NormFinder, and BestKeeper to compare expression stability, we found that GAPDH and HPRT1 were the most unstable reference genes among all those tested. Stability analysis yielded very similar results when geNorm or BestKeeper was used but not when NormFinder was used. Specifically, geNorm and BestKeeper identified RPL13 and RPLP0 as the most stable genes under 20 μM Lip-1 treatment, whereas RPL13, EEF1A1, and TUBB were the most stable under 10 μM Lip-1 treatment. TUBB and EEF1A1 were the most stable genes in both treatment groups according to the results obtained using NormFinder. An assumed most stable gene was incorporated into each software to validate the accuracy. The results suggest that NormFinder is not an appropriate algorithm for this study.

CONCLUSIONS

Stable reference genes were recognized using geNorm and BestKeeper but not NormFinder. Overall, RPL13 and RPLP0 were the most stable reference genes under 20 μM Lip-1 treatment, whereas RPL13, EEF1A1, and TUBB were the most stable genes under 10 μM Lip-1 treatment.

Association between excessive screen time and school-level proportion of no family rules among elementary school children in Japan: a multilevel analysis

BACKGROUND

Excessive screen time (ST) in children is a global concern. We assessed the association between individual- and school-level factors and excessive ST in Japanese children using a multilevel analysis.

METHODS

A school-based cross-sectional study was conducted in Toyama, Japan in 2018. From 110 elementary schools in Toyama Prefecture, 13,413 children in the 4th-6th grades (boys, 50.9%; mean, 10.5 years old) participated. We assessed lifestyle, recreational ST (not for study use), psychological status, and school and family environment including family rules. We defined ≥3 hours ST as excessive. We calculated the school-level proportions of no family rules and divided them into four categories (<20%, 20% to <30%, 30% to <40%, and ≥40%). A modified multilevel Poisson regression analysis was performed.

RESULTS

In total, 12,611 children were included in the analysis (94.0%). The average school-level proportion of those with no family rules was 32.1% (SD = 9.6). The prevalence of excessive ST was 29.9% (34.9% in boys; 24.8% in girls). The regression analysis showed that excessive ST was significantly associated with both individual-level factors, such as boys (adjusted prevalence ratio (aPR); 1.39), older grades (aPR; 1.18 for 5th grades and 1.28 for 6th grades), late wakeup (aPR; 1.13), physical inactivity (aPR; 1.18 for not so much and 1.31 for rarely), late bedtime (aPR; 1.43 for 10 to 11 p.m. and 1.76 for ≥11 p.m.), frequent irritability (aPR; 1.24 for sometimes and 1.46 for often), feelings of school avoidance (aPR; 1.17 for sometimes and 1.22 for often), infrequent child-parental interaction (aPR; 1.16 for rare and 1.21 for none), no family rules (aPR; 1.56), smartphone ownership (aPR; 1.18), and the school-level proportion of no family rules (aPR; 1.20 for 20% to <30%, 1.29 for 30% to <40%, and 1.43 for ≥40%, setting <20% as reference).

CONCLUSION

Besides individual factors, a higher school-level proportion of no family rules seemed influential on excessive ST. Increasing the number of households with family rules and addressing individual factors, could be deterrents against excessive ST in children.

Considerations for future modification of The Association for the Study of Reproductive Biology embryo grading system incorporating time-lapse observations

The Association for the Study of Reproductive Biology (ASEBIR) Interest Group in Embryology (in Spanish 'Grupo de Interés de Embriología') reviewed key morphokinetic parameters to assess the contribution of time-lapse technology (TLT) to the ASEBIR grading system. Embryo grading based on morphological characteristics is the most widely used method in human assisted reproduction laboratories. The introduction and implementation of TLT has provided a large amount of information that can be used as a complementary tool for morphological embryo evaluation and selection. As part of IVF treatments, embryologists grade embryos to decide which embryos to transfer or freeze. At the present, the embryo grading system developed by ASEBIR does not consider dynamic events observed through TLT. Laboratories that are using TLT consider those parameters as complementary data for embryo selection. The aim of this review was to evaluate review time-specific morphological changes during embryo development that are not included in the ASEBIR scoring system, and to consider them as candidates to add to the scoring system.

Added value of an artificial intelligence algorithm in reducing the number of missed incidental acute pulmonary embolism in routine portal venous phase chest CT

OBJECTIVES

The purpose of this study was to evaluate the incremental value of artificial intelligence (AI) compared to the diagnostic accuracy of radiologists alone in detecting incidental acute pulmonary embolism (PE) on routine portal venous contrast-enhanced chest computed tomography (CT).

METHODS

CTs of 3089 consecutive patients referred to the radiology department for a routine contrast-enhanced chest CT between 27-5-2020 and 31-12-2020, were retrospectively analysed by a CE-certified and FDA-approved AI algorithm. The diagnostic performance of the AI was compared to the initial report. To determine the reference standard, discordant findings were independently evaluated by two readers. In case of disagreement, another experienced cardiothoracic radiologist with knowledge of the initial report and the AI output adjudicated.

RESULTS

The prevalence of acute incidental PE in the reference standard was 2.2% (67 of 3089 patients). In 25 cases, AI detected initially unreported PE. This included three cases concerning central/lobar PE. Sensitivity of the AI algorithm was significantly higher than the outcome of the initial report (respectively 95.5% vs. 62.7%, p < 0.001), whereas specificity was very high for both (respectively 99.6% vs 99.9%, p = 0.012). The AI algorithm only showed a slightly higher amount of false-positive findings (11 vs. 2), resulting in a significantly lower PPV (85.3% vs. 95.5%, p = 0.047).

CONCLUSION

The AI algorithm showed high diagnostic accuracy in diagnosing incidental PE, detecting an additional 25 cases of initially unreported PE, accounting for 37.3% of all positive cases.

CLINICAL RELEVANCE STATEMENT

Radiologist support from AI algorithms in daily practice can prevent missed incidental acute PE on routine chest CT, without a high burden of false-positive cases.

KEY POINTS

• Incidental pulmonary embolism is often missed by radiologists in non-diagnostic scans with suboptimal contrast opacification within the pulmonary trunk. • An artificial intelligence algorithm showed higher sensitivity detecting incidental pulmonary embolism on routine portal venous chest CT compared to the initial report. • Implementation of artificial intelligence support in routine daily practice will reduce the number of missed incidental pulmonary embolism.

Artificial intelligence predicts direct-acting antivirals failure among hepatitis C virus patients: A nationwide hepatitis C virus registry program

BACKGROUND/AIMS

Despite the high efficacy of direct-acting antivirals (DAAs), approximately 1-3% of hepatitis C virus (HCV) patients fail to achieve a sustained virological response. We conducted a nationwide study to investigate risk factors associated with DAA treatment failure. Machine-learning algorithms have been applied to discriminate subjects who may fail to respond to DAA therapy.

METHODS

We analyzed the Taiwan HCV Registry Program database to explore predictors of DAA failure in HCV patients. Fifty-five host and virological features were assessed using multivariate logistic regression, decision tree, random forest, eXtreme Gradient Boosting (XGBoost), and artificial neural network. The primary outcome was undetectable HCV RNA at 12 weeks after the end of treatment.

RESULTS

The training (n=23,955) and validation (n=10,346) datasets had similar baseline demographics, with an overall DAA failure rate of 1.6% (n=538). Multivariate logistic regression analysis revealed that liver cirrhosis, hepatocellular carcinoma, poor DAA adherence, and higher hemoglobin A1c were significantly associated with virological failure. XGBoost outperformed the other algorithms and logistic regression models, with an area under the receiver operating characteristic curve of 1.000 in the training dataset and 0.803 in the validation dataset. The top five predictors of treatment failure were HCV RNA, body mass index, α-fetoprotein, platelets, and FIB-4 index. The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the XGBoost model (cutoff value=0.5) were 99.5%, 69.7%, 99.9%, 97.4%, and 99.5%, respectively, for the entire dataset.

CONCLUSION

Machine learning algorithms effectively provide risk stratification for DAA failure and additional information on the factors associated with DAA failure.

Development of an algorithm for finding pertussis episodes in a population-based electronic health record database

While tetanus-diphtheria-acellular pertussis (Tdap) vaccines for adolescents and adults were licensed in 2005 and immunization strategies proposed, the burden of pertussis in this population remains under-recognized mainly due to atypical disease presentation, undermining efforts to optimize protection through vaccination. We developed a machine learning algorithm to identify undiagnosed/misdiagnosed pertussis episodes in patients diagnosed with acute respiratory disease (ARD) using signs, diseases and symptoms from clinician notes and demographic information within electronic health-care records (Optum Humedica repository [2007-2019]). We used two patient cohorts aged ≥11 years to develop the model: a positive pertussis cohort (4,515 episodes in 4,316 patients) and a negative pertussis (ARD) cohort (4,573,445 episodes and patients), defined using ICD 9/10 codes. To improve contrast between positive pertussis and negative pertussis (ARD) episodes, only episodes with ≥7 symptoms were selected. LightGBM was used as the machine learning model for pertussis episode identification. Model validity was determined using laboratory-confirmed pertussis positive and negative cohorts. Model explainability was obtained using the Shapley additive explanations method. The predictive performance was as follows: area under the precision-recall curve, 0.24 (SD, 7 × 10^-3); recall, 0.72 (SD, 4 × 10^-3); precision, 0.012 (SD, 1 × 10^-3); and specificity, 0.94 (SD, 7 × 10^-3). The model applied to laboratory-confirmed positive and negative pertussis episodes had a specificity of 0.846. Predictive probability for pertussis increased with presence of whooping cough, whoop, and post-tussive vomiting in clinician notes, but decreased with gastrointestinal bleeding, sepsis, pulmonary symptoms, and fever. In conclusion, machine learning can help identify pertussis episodes among those diagnosed with ARD.

CT-based deep learning model for predicting hospital discharge outcome in spontaneous intracerebral hemorrhage

OBJECTIVES

To predict the functional outcome of patients with intracerebral hemorrhage (ICH) using deep learning models based on computed tomography (CT) images.

METHODS

A retrospective, bi-center study of ICH patients was conducted. Firstly, a custom 3D convolutional model was built for predicting the functional outcome of ICH patients based on CT scans from randomly selected ICH patients in H training dataset collected from H hospital. Secondly, clinical data and radiological features were collected at admission and the Extreme Gradient Boosting (XGBoost) algorithm was used to establish a second model, named the XGBoost model. Finally, the Convolution model and XGBoost model were fused to build the third "Fusion model." Favorable outcome was defined as modified Rankin Scale score of 0-3 at discharge. The prognostic predictive accuracy of the three models was evaluated using an H test dataset and an external Y dataset, and compared with the performance of ICH score and ICH grading scale (ICH-GS).

RESULTS

A total of 604 patients with ICH were included in this study, of which 450 patients were in the H training dataset, 50 patients in the H test dataset, and 104 patients in the Y dataset. In the Y dataset, the areas under the curve (AUCs) of the Convolution model, XGBoost model, and Fusion model were 0.829, 0.871, and 0.905, respectively. The Fusion model prognostic performance exceeded that of ICH score and ICH-GS (p = 0.043 and p = 0.045, respectively).

CONCLUSIONS

Deep learning models have good accuracy for predicting functional outcome of patients with spontaneous intracerebral hemorrhage.

CLINICAL RELEVANCE STATEMENT

The proposed deep learning Fusion model may assist clinicians in predicting functional outcome and developing treatment strategies, thereby improving the survival and quality of life of patients with spontaneous intracerebral hemorrhage.

KEY POINTS

• Integrating clinical presentations, CT images, and radiological features to establish deep learning model for functional outcome prediction of patients with intracerebral hemorrhage. • Deep learning applied to CT images provides great help in prognosing functional outcome of intracerebral hemorrhage patients. • The developed deep learning model performs better than clinical prognostic scores in predicting functional outcome of patients with intracerebral hemorrhage.

Identifying pregnancy episodes and estimating the last menstrual period using an administrative database in Korea: an application to patients with systemic lupus erythematosus

OBJECTIVES

This study developed an algorithm for identifying pregnancy episodes and estimating the last menstrual period (LMP) in an administrative claims database and applied it to investigate the use of pregnancy-incompatible immunosuppressants among pregnant women with systemic lupus erythematosus (SLE).

METHODS

An algorithm was developed and applied to a nationwide claims database in Korea. Pregnancy episodes were identified using a hierarchy of pregnancy outcomes and clinically plausible periods for subsequent episodes. The LMP was estimated using preterm delivery, sonography, and abortion procedure codes. Otherwise, outcome-specific estimates were applied, assigning a fixed gestational age to the corresponding pregnancy outcome. The algorithm was used to examine the prevalence of pregnancies and utilization of pregnancy-incompatible immunosuppressants (cyclophosphamide [CYC]/mycophenolate mofetil [MMF]/methotrexate [MTX]) and non-steroidal anti-inflammatory drugs (NSAIDs) during pregnancy in SLE patients.

RESULTS

The pregnancy outcomes identified in SLE patients included live births (67%), stillbirths (2%), and abortions (31%). The LMP was mostly estimated with outcome-specific estimates for full-term births (92.3%) and using sonography procedure codes (54.7%) and preterm delivery diagnosis codes (37.9%) for preterm births. The use of CYC/MMF/MTX decreased from 7.6% during preconception to 0.2% at the end of pregnancy. CYC/MMF/MTX use was observed in 3.6% of women within 3 months preconception and 2.5% during 0-7 weeks of pregnancy.

CONCLUSIONS

This study presents the first pregnancy algorithm using a Korean administrative claims database. Although further validation is necessary, this study provides a foundation for evaluating the safety of medications during pregnancy using secondary databases in Korea, especially for rare diseases.

Enhancing artificial intelligence-doctor collaboration for computer-aided diagnosis in colonoscopy through improved digital literacy

Establishing appropriate trust and maintaining a balanced reliance on digital resources are vital for accurate optical diagnoses and effective integration of computer-aided diagnosis (CADx) in colonoscopy. Active learning using diverse polyp image datasets can help in developing precise CADx systems. Enhancing doctors' digital literacy and interpreting their results is crucial. Explainable artificial intelligence (AI) addresses opacity, and textual descriptions, along with AI-generated content, deepen the interpretability of AI-based findings by doctors. AI conveying uncertainties and decision confidence aids doctors' acceptance of results. Optimal AI-doctor collaboration requires improving algorithm performance, transparency, addressing uncertainties, and enhancing doctors' optical diagnostic skills.

Added value of clinical decision rules for the management of enhancing breast MRI lesions: A systematic comparison of the Kaiser score and the Göttingen score

PURPOSE

We investigated the added value of two internationally used clinical decision rules in the management of enhancing lesions on breast MRI.

METHODS

This retrospective, institutional review board approved study included consecutive patients from two different populations. Patients received breast MRI according to the recommendations of the European Society of Breast Imaging (EUSOBI). Initially, all examinations were assessed by expert readers without using clinical decision rules. All lesions rated as category 4 or 5 according to the Breast Imaging Reporting and Data System were histologically confirmed. These lesions were re-evaluated by an expert reader blinded to the histology. He assigned each lesion a Göttingen score (GS) and a Kaiser score (KS) on different occasions. To provide an estimate on inter-reader agreement, a second fellowship-trained reader assessed a subset of these lesions. Subgroup analyses based on lesion type (mass vs. non-mass), size (>1 cm vs. ≤ 1 cm), menopausal status, and significant background parenchymal enhancement were conducted. The areas under the ROC curves (AUCs) for the GS and KS were compared, and the potential to avoid unnecessary biopsies was determined according to previously established cutoffs (KS > 4, GS > 3) RESULTS: 527 lesions in 506 patients were included (mean age: 51.8 years, inter-quartile-range: 43.0-61.0 years). 131/527 lesions were malignant (24.9 %; 95 %-confidence-interval: 21.3-28.8). In all subgroups, the AUCs of the KS (median = 0.91) were higher than those of the GS (median = 0.83). Except for "premenopausal patients" (p = 0.057), these differences were statistically significant (p ≤ 0.01). Kappa agreement was higher for the KS (0.922) than for the GS (0.358).

CONCLUSION

Both the KS and the GS provided added value for the management of enhancing lesions on breast MRI. The KS was superior to the GS in terms of avoiding unnecessary biopsies and showed superior inter-reader agreement; therefore, it may be regarded as the clinical decision rule of choice.

Spacetimematter of aging - The material temporalities of later life

Material gerontology poses the question of how aging processes are co-constituted in relation to different forms of (human and non-human) materiality. This paper makes a novel contribution by asking when aging processes are co-constituted and how these temporalities of aging are entangled with different forms of materiality. In this paper, we explore the entanglements of temporality and materiality in shaping later life by framing them as spacetimematters (Barad, 2013). By drawing on empirical examples from data from a qualitative case study in a long-term care (LTC) facility, we ask how the entanglement of materiality and temporality of a fall-detection sensor co-constitutes aging. We focus on two types of material temporality that came to matter in age-boundary-making practices at this site: the material temporality of a technology-in-training and the material temporality of (false) alarms. Both are interwoven, produced and reproduced through spacetimematterings that established age-boundaries. Against the backdrop of these findings, we propose to understand age(ing) as a situated, distributed, more-than-human process of practices: It emerges in an assemblage of technological innovation discourses, problematizations of demographic change, digitized and analog practices of care and caring, bodily functioning, daily routines, institutionalized spaces and much more. Finally, we discuss the role power plays in those spacetimematterings of aging and conclude with a research outlook for material gerontology.

Artificial intelligence: Machine learning approach for screening large database and drug discovery

Recent research in drug discovery dealing with many faces difficulties, including development of new drugs during disease outbreak and drug resistance due to rapidly accumulating mutations. Virtual screening is the most widely used method in computer aided drug discovery. It has a prominent ability in screening drug targets from large molecular databases. Recently, a number of web servers have developed for quickly screening publicly accessible chemical databases. In a nutshell, deep learning algorithms and artificial neural networks have modernised the field. Several drug discovery processes have used machine learning and deep learning algorithms, including peptide synthesis, structure-based virtual screening, ligand-based virtual screening, toxicity prediction, drug monitoring and release, pharmacophore modelling, quantitative structure-activity relationship, drug repositioning, polypharmacology, and physiochemical activity. Although there are presently a wide variety of data-driven AI/ML tools available, the majority of these tools have, up to this point, been developed in the context of non-communicable diseases like cancer, and a number of obstacles have prevented the translation of these tools to the discovery of treatments against infectious diseases. In this review various aspects of AI and ML in virtual screening of large databases were discussed. Here, with an emphasis on antivirals as well as other disease, offers a perspective on the advantages, drawbacks, and hazards of AI/ML techniques in the search for innovative treatments.

Impact of adherence to disease-modifying drugs in multiple sclerosis: A study on Italian real-world data

BACKGROUND

Multiple Sclerosis (MS) is a chronic inflammatory disease of the central nervous system requiring complex diagnostic and therapeutic management. Treatment with Disease Modifying Drugs (DMDs) is aimed at reducing relapse rate and disease disability. Few real-world, population-based data are available on the impact of adherence on relapse rate. The objective of this study was to assess the impact of adherence to DMDs on relapses in a real-world Italian setting.

METHODS

Population-based cohort study. People with MS (PwMS) older than 18 years and residing in the Emilia-Romagna region, Northern Italy, were identified through administrative databases using a validated algorithm. A Cox regression model with a time-varying exposure was performed to assess the association between level of adherence to DMDs and relapses over a 5-year period.

RESULTS

A total of 2,528 PwMS receiving a first prescription of DMDs between 2015 and 2019 were included (average age of 42, two-thirds female). Highly adherent PwMS had a 25 % lower hazard of experiencing moderate or severe relapses than non-adherent PwMS (Hazard Ratio 0.75, 95 % CI 0.58 to 0.98), after adjusting for age and sex. Several sensitivity analyses supported the main result.

CONCLUSION

The results of our study support the hypothesis that a high level of DMD adherence in MS is associated with a lower risk of moderate or severe relapse. Therefore, choosing the DMD with which to start drug treatment and recommending adherence to treatment appear to be crucial aspects involving both physicians and patients.

Comparing methods for constructing and representing human pangenome graphs

BACKGROUND

As a single reference genome cannot possibly represent all the variation present across human individuals, pangenome graphs have been introduced to incorporate population diversity within a wide range of genomic analyses. Several data structures have been proposed for representing collections of genomes as pangenomes, in particular graphs.

RESULTS

In this work, we collect all publicly available high-quality human haplotypes and construct the largest human pangenome graphs to date, incorporating 52 individuals in addition to two synthetic references (CHM13 and GRCh38). We build variation graphs and de Bruijn graphs of this collection using five of the state-of-the-art tools: Bifrost, mdbg, Minigraph, Minigraph-Cactus and pggb. We examine differences in the way each of these tools represents variations between input sequences, both in terms of overall graph structure and representation of specific genetic loci.

CONCLUSION

This work sheds light on key differences between pangenome graph representations, informing end-users on how to select the most appropriate graph type for their application.

Estimation of the availability of iron in the school meals of Municipal Centers for Early Childhood Education of a capital city in northeastern Brazil

The final stage of Fe deficiency is Fe deficiency anaemia, with repercussions for human health, especially in children under 5 years of age. Studies conducted in Brazilian public daycare centres show high prevalence of anaemia. The present study aims to evaluate the availability of Fe in the meals of the Municipal Centers of Early Childhood Education in Maceió. The experimental design comprises selection of algorithms, menu evaluation, calculation of the estimates, comparison between the estimates obtained and the recommendations, and analysis of correlation between meal constituents, and of the concordance between the absorbable Fe estimates. Four algorithms were selected and a monthly menu consisting of 22 d. The correlation analysis showed a moderate positive correlation to animal tissue (AT) v. non-heme iron (r = 0·42; P = 0·04), and negative to AT v. Ca (r = -0·54; P = 0·09) and Ca v. phytates (r = -0·46, P = 0·03). Estimates of absorbable Fe ranged from 0·23 to 0·44 mg/d. The amount of Fe available, unlike the total amount of Fe offered, does not meet the nutritional recommendations on most school days. The Bland-Altman analysis indicated that the Monsen and Balinfty and Rickard et al. showed greater agreement. The results confirm the need to adopt strategies to increase the availability of Fe in school meals.