Infertility has emerged as a significant public health concern, with assisted reproductive technology (ART) is a last-resort treatment option. However, ART's efficacy is limited by significant financial cost and physical discomfort. The aim of this study is to build Machine learning (ML) decision-support models to predict the optimal range of embryo numbers to transfer, using data from infertile couples identified through literature reviews. Binary classification models were developed to classify cases into two groups: those transferring two or fewer embryos and those transferring three or four. Four popular ML algorithms were used, including random forest (RF), logistic regression (LR), support vector machine (SVM), and artificial neural network (ANN), considering seven criteria: the woman's age, sperm origin, the developmental qualities of four potential embryos, infertility duration, assessment of the woman, morphological qualities of the four best embryos on the day of transfer, and number of oocytes extracted. The stratified 3-fold cross-validation results show that the SVM model obtained the highest average accuracy (95.83%) and demonstrated the best overall performance, closely followed by the ANN and LR models with an average accuracy equal to 91.67%. The RF model achieved a slightly lower average accuracy (88.89%), which demonstrated the lowest variability. Testing on a new dataset revealed all models performed well, with ANN and SVM models classified all test set instances correctly, while the RF and LR models achieved 91.68% accuracy. These results highlight the superior generalization and effectiveness of the ANN and SVM models in guiding ART decisions.

• Klíčová slova
• Artificial neural network, assisted reproductive technology, embryo transfer, infertility, multi-criteria decision aiding, number of embryos,
• MeSH
• asistovaná reprodukce * MeSH
• dospělí MeSH
• infertilita * terapie   MeSH
• lidé MeSH
• neuronové sítě MeSH
• přenos embrya * MeSH
• strojové učení * MeSH
• support vector machine MeSH
• těhotenství MeSH
• Check Tag
• dospělí MeSH
• lidé MeSH
• mužské pohlaví MeSH
• těhotenství MeSH
• ženské pohlaví MeSH
• Publikační typ
• časopisecké články MeSH

Spectroscopic data often contain artifacts or noise related to the sample characteristics, instrumental variations, or experimental design flaws. Therefore, classifying the raw data is not recommended and might lead to biased results. Nevertheless, most issues may be addressed through appropriate data pre-processing. Effective pre-processing is particularly crucial in critical applications like liquid biopsy for disease detection, where even minor performance improvements may impact patient outcomes. Unfortunately, there is no consensus regarding optimal pre-processing, complicating cross-study comparisons. This study presents a comprehensive evaluation of various pre-processing methods and their combinations to assess their influence on classification results. The goal was to identify whether some pre-processing methods are associated with higher classification outcomes and find an optimal strategy for the given data. Data from Raman optical activity and infrared and Raman spectroscopy were processed, applying tens of thousands of possible pre-processing pipelines. The resulting data were classified using three algorithms to distinguish between subjects with liver cirrhosis and those who had developed hepatocellular carcinoma. Results highlighted that some specific pre-processing methods often ranked among the best classification results, such as the Rolling Ball for correcting the baseline of Raman spectra or the Doubly Reweighted Penalized Least Squares and Mixture model in the case of Raman optical activity. On the other hand, the selection of filtering and/or normalization approach usually did not have a significant impact. Nonetheless, the pre-processing of top-scoring pipelines also depended on the classifier utilized. The best pipelines yielded an AUROC of 0.775-0.823, varying with the evaluated spectroscopic data and classifier.

• Klíčová slova
• Chiroptical spectroscopy, Classification, Data pre-processing, Diagnostics, Liquid biopsy, Machine learning, Vibrational spectroscopy,
• MeSH
• algoritmy MeSH
• hepatocelulární karcinom * diagnóza   patologie   MeSH
• jaterní cirhóza diagnóza   patologie   MeSH
• lidé MeSH
• metoda nejmenších čtverců MeSH
• nádory jater * diagnóza   patologie   MeSH
• Ramanova spektroskopie * metody   MeSH
• spektrofotometrie infračervená metody   MeSH
• tekutá biopsie metody   MeSH
• Check Tag
• lidé MeSH
• Publikační typ
• časopisecké články MeSH

BACKGROUND AND OBJECTIVE: Patient-ventilator asynchronies (PVA) are associated with ventilator-induced lung injury and increased mortality. Current detection methods rely on static thresholds, extensive preprocessing, or proprietary ventilator data. This study aimed to develop and validate a fully online, real-time system that detects and classifies PVAs directly from ventilator screen data while alerting clinicians based on severity. METHODS: The SmartAlert system was developed using ventilator screen recordings from ICU patients. It extracts pressure and flow waveforms from video recordings, converts them into time-series data, and employs deep neural networks to classify asynchronies and assign alarm levels from no urgency to most urgent. A dataset of 381,280 double-breath units was independently annotated by two expert intensivists. Two deep learning models were trained: one for alarm prediction and another for asynchrony classification (ineffective triggering, double cycling, high inspiratory effort, no asynchrony). Performance was evaluated using accuracy, sensitivity, specificity, and AUC-ROC, compared to expert consensus. RESULTS: SmartAlert demonstrated strong performance for alarm level prediction (overall accuracy: 83.8 %, weighted AUC-ROC: 0.943 [95 % CI: 0.941-0.945]) and PVA classification (weighted accuracy: 89.3 %, weighted AUC-ROC: 0.951 [95 % CI: 0.950-0.953]). It showed high specificity for urgent alarms (99.9 % for level 3) and PVA types (98.5 % for ineffective triggering, 96.9 % for double cycling, 94.8 % for high inspiratory effort). CONCLUSIONS: We developed and internally validated SmartAlert, an automated system that detects PVAs, classifies severity, and alerts clinicians in real time. Its potential to reduce alarm fatigue, optimize ventilator settings, and improve patient outcomes remains to be tested in clinical trials.

• Klíčová slova
• Deep neural networks, Mechanical ventilation, Patient-ventilator asynchrony, Real-time monitoring,
• MeSH
• asynchronie mezi pacientem a ventilátorem MeSH
• deep learning MeSH
• jednotky intenzivní péče * MeSH
• klinické alarmy MeSH
• lidé MeSH
• mechanické ventilátory * MeSH
• neuronové sítě MeSH
• reprodukovatelnost výsledků MeSH
• ROC křivka MeSH
• strojové učení * MeSH
• umělé dýchání * MeSH
• Check Tag
• lidé MeSH
• Publikační typ
• časopisecké články MeSH

OBJECTIVES: This randomized controlled trial aimed to evaluate the impact of artificial intelligence (AI) assistance on dentists' diagnostic accuracy, confidence, and treatment decisions when detecting periapical radiolucencies (PRs) on panoramic radiographs. We specifically investigated whether AI support influenced diagnostic performance across different levels of clinical experience. METHODS: Thirty dentists with varying levels of experience evaluated 50 panoramic radiographs for the presence or absence of PRs, with and without the aid of AI, using a cross-over design. Diagnostic performance metrics, confidence scores, and clinical decision choices were analyzed. CBCT scans served as the reference standard. Outcomes included sensitivity, specificity, positive and negative predictive values, overall diagnostic accuracy, and area under the ROC and AFROC curves. Statistical analyses were conducted using mixed-effects regression models. RESULTS: AI assistance significantly improved overall diagnostic accuracy (91.6 % unaided vs. 93.3 % AI-aided; p < 0.001), mainly by reducing false positive diagnoses (false positive rate: 4.3 % unaided vs. 2.0 % AI-aided). Sensitivity remained stable (46.0 % unaided vs. 45.8 % AI-aided). Junior dentists showed the greatest improvements in performance and confidence. AI support shifted treatment decisions toward more conservative approaches. CONCLUSIONS: AI assistance modestly enhanced dentists' diagnostic accuracy for detecting periapical radiolucencies, primarily by decreasing false positive diagnoses. Junior dentists benefited most from AI support. Integration of AI in diagnostic workflows may reduce overtreatment and enhance diagnostic consistency, especially among less experienced clinicians. CLINICAL SIGNIFICANCE: The integration of AI support in dental diagnostics reduced false positive diagnoses and supported more conservative treatment decisions, particularly benefiting less experienced clinicians. These findings suggest that AI assistance can enhance diagnostic consistency and reduce overtreatment in clinical dental practice.

• Klíčová slova
• Artificial intelligence, Dentistry, Diagnostic accuracy, Panoramic radiography, Periapical radiolucency, Randomized controlled trial,
• MeSH
• dospělí MeSH
• falešně pozitivní reakce MeSH
• klinické kompetence MeSH
• klinické křížové studie MeSH
• klinické rozhodování MeSH
• lidé středního věku MeSH
• lidé MeSH
• periapikální nemoci * diagnostické zobrazování   MeSH
• počítačová tomografie s kuželovým svazkem MeSH
• rentgendiagnostika panoramatická MeSH
• senzitivita a specificita MeSH
• umělá inteligence * MeSH
• Check Tag
• dospělí MeSH
• lidé středního věku MeSH
• lidé MeSH
• mužské pohlaví MeSH
• ženské pohlaví MeSH
• Publikační typ
• časopisecké články MeSH
• randomizované kontrolované studie MeSH

OBJECTIVE: Artificial intelligence (AI) is increasingly used in dental research for diagnosis, treatment planning, and disease prediction. However, many dental AI studies lack methodological rigor, transparency, or reproducibility, and no dedicated peer-review guidance exists for this field. METHODS: Editors and reviewers from the ITU/WHO/WIPO AI for Health - Dentistry group participated in a structured survey and group discussions to identify key elements for reviewing AI dental research. A draft of the recommendations was circulated for feedback and consensus. RESULTS: The consensus from editors and reviewers identified four key indicators of high-quality AI dental research: (1) relevance to a real clinical or methodological problem, (2) robust and transparent methodology, (3) reproducibility through data/code availability or functional demos, and (4) adherence to ethical and responsible reporting practices. Common reasons for rejection included lack of novelty, poor methodology, limited external testing, and overstated claims. Four essential checks were proposed to support peer review: the study should address a meaningful clinical question, follow appropriate reporting guidelines (e.g., DENTAL-AI, STARD-AI), clearly describe reproducible methods, and use precise, justified, and clinically relevant wording. CONCLUSION: Editors and reviewers play a critical role in improving the quality of AI research in dentistry. This guidance aims to support more robust peer review and contribute to the development of reliable, clinically relevant, and ethically sound AI applications in dentistry.

• Klíčová slova
• Artificial intelligence, Deep learning, Dentistry, Machine learning, Peer-review,
• MeSH
• lidé MeSH
• posudkové řízení ve výzkumu * normy   MeSH
• posudkové řízení MeSH
• reprodukovatelnost výsledků MeSH
• stomatologický výzkum * normy   MeSH
• umělá inteligence * MeSH
• výzkumný projekt normy   MeSH
• Check Tag
• lidé MeSH
• Publikační typ
• časopisecké články MeSH

A recurring discrepancy in attitudes toward decisions made by human versus artificial agents, termed the Human-Robot moral judgment asymmetry, has been documented in moral psychology of AI. Across a wide range of contexts, AI agents are subject to greater moral scrutiny than humans for the same actions and decisions. In eight experiments (total N = 5837), we investigated whether the asymmetry effect arises in end-of-life care contexts and explored the mechanisms underlying this effect. Our studies documented reduced approval of an AI doctor's decision to withdraw life support relative to a human doctor (Studies 1a and 1b). This effect persisted regardless of whether the AI assumed a recommender role or made the final medical decision (Studies 2a and 2b and 3), but, importantly, disappeared under two conditions: when doctors kept on rather than withdraw life support (Studies 1a, 1b and 3), and when they carried out active euthanasia (e.g., providing a lethal injection or removing a respirator on the patient's demand) rather than passive euthanasia (Study 4). These findings highlight two contextual factors-the level of automation and the patient's autonomy-that influence the presence of the asymmetry effect, neither of which is not predicted by existing theories. Finally, we found that the asymmetry effect was partly explained by perceptions of AI incompetence (Study 5) and limited explainability (Study 6). As the role of AI in medicine continues to expand, our findings help to outline the conditions under which stakeholders disfavor AI over human doctors in clinical settings.

• Klíčová slova
• AI ethics, Moral judgment, Moral psychology of AI, Moral psychology of robotics, Passive euthanasia,
• MeSH
• dospělí MeSH
• eutanazie * psychologie   MeSH
• lidé středního věku MeSH
• lidé MeSH
• mínění MeSH
• mladý dospělý MeSH
• mravy * MeSH
• rozhodování * MeSH
• umělá inteligence * MeSH
• Check Tag
• dospělí MeSH
• lidé středního věku MeSH
• lidé MeSH
• mladý dospělý MeSH
• mužské pohlaví MeSH
• ženské pohlaví MeSH
• Publikační typ
• časopisecké články MeSH
• přehledy MeSH

Preparative chromatography for purification of sugars has found numerous applications in food processing and biochemical industries. Parameter estimation leading to robust modelling and model-based design are essential for transferring this technology into industrial practice. This study examines the Equilibrium Dispersive model with a non-linear isotherm and the formulation of apparent dispersion based on the Bodenstein number. A parameter estimation workflow is proposed, incorporating chromatography-specific algorithmic data preprocessing and a curve uncertainty scoring system, enabling the simultaneous utilization of data from pulse-feed experiments conducted under varying conditions. A bilevel optimization scheme is introduced, leading to increased performance and robustness. The introduction of parameter bounding and initialization eliminates arbitrariness in the process. Experiments on the chromatographic separation of a d-glucose and sucrose mixture, performed under different flow rates and feed loads using an anion-exchange resin, were conducted. Lab-scale experiments were used for parameter estimation, supported by subsequent identifiability and sensitivity analyses. Scaling-up predictions of the calibrated model were evaluated by experimental data from a 2-meter-long pilot-scale column. The results demonstrate the benefits of the proposed modeling and parameter estimation framework, as well as the sufficient predictive accuracy of the calibrated model under the conditions of scaled-up flow rates and column dimensions.

• Klíčová slova
• Equilibrium dispersive model, Parameter estimation and bilevel optimization, Preparative chromatography, Process scale-up modelling and pilot plant evaluation, Sugars separation,
• MeSH
• algoritmy MeSH
• chemické modely * MeSH
• chromatografie iontoměničová metody   MeSH
• glukosa * izolace a purifikace   chemie   analýza   MeSH
• nelineární dynamika MeSH
• sacharosa * izolace a purifikace   chemie   MeSH
• Publikační typ
• časopisecké články MeSH
• Názvy látek
• glukosa * MeSH
• sacharosa * MeSH

Selection of the optimal makeup solvent composition is critical for achieving sensitive and reproducible ionization in supercritical fluid chromatography-mass spectrometry (SFC-MS). This study investigated the ionization processes in a spray-based ionization source called UniSpray (US), by an artificial neural network driven approach, emphasizing the effect of makeup solvent composition. A set of compounds with different physicochemical properties was analyzed using a generic SFC method and 24 makeup solvents. Artificial neural networks were used to correlate molecular descriptors with MS responses and to identify key analyte properties affecting ionization. Statistical analysis of this extensive dataset revealed significant differences in ionization efficiency compared to electrospray ionization (ESI), depending on makeup solvent composition and analyte properties. While US outperformed ESI for 82 % of compounds, certain analytes, including basic beta-blockers, fluorine-substituted compounds, and small lipophilic molecules, benefited from ESI. Optimized makeup solvent compositions differed notably between ESI and US. For example, ethanol and isopropanol were recommended for US+ but not for ESI+. The use of water and ammonia also affected MS responses differently between sources and ionization modes, with optimal concentrations varying depending on the analyte and organic modifier of the SFC mobile phase. This study highlights key differences between SFC-ESI-MS and SFC-US-MS ionization efficiency and demonstrates the utility of data-driven methodologies for faster and more efficient method development.

• Klíčová slova
• Artificial neural networks, Chemometrics, Electrospray, Mass spectrometry, Supercritical fluid chromatography, UniSpray,
• MeSH
• hmotnostní spektrometrie s elektrosprejovou ionizací * metody   MeSH
• hmotnostní spektrometrie * metody   MeSH
• neuronové sítě MeSH
• rozpouštědla chemie   MeSH
• superkritická fluidní chromatografie * metody   MeSH
• umělá inteligence * MeSH
• Publikační typ
• časopisecké články MeSH
• Názvy látek
• rozpouštědla MeSH

BACKGROUND AND OBJECTIVE: Digital Holographic Microscopy provides a new kind of quantitative image data about live cells' in vitro activities. Apart from non-invasive and staining-free imaging, it offers topological weighting of cell mass. This led us to develop a particular tool for assessing cell mass dynamics. METHODS: Programming language Python and a training set of time-lapse images of adherent HT-1080 cells derived from human fibrosarcoma taken with dry objective 40x/0.95 at 30-second intervals were used to create the Analytical Image Differencing (AID) method. RESULTS: The AID makes the best of these new data by evaluating the difference between the chosen two quantitative phase images from the time-lapse series. The contribution of the method is demonstrated on hiQPI (Holographic Incoherent-light-source Quantitative Phase Imaging) image data taken with a Q-phase microscope. The analysis outputs are graphical and complemented with numerical data. To underscore the significance of the Analytical Image Differencing (AID) method, an initial pilot experiment was conducted to show the available analyses of sequential overlapping images capturing the movement of cancer cells. Notably, besides defining changes in areas used by the cell (newly or steadily occupied or better abandoned) it is an introduction to the zero-line concept, which denotes spots of tranquility among continuously moving surroundings. CONCLUSIONS: The measurement of zero-line length has emerged as a novel biomarker for characterizing cell mass transfer. The sensitivity of phase change measurements is demonstrated. The noise quality of input images obtained with incoherent (hiQPI) and coherent (QPI) methods is compared. The resulting effect on the AID method output is also shown. The findings of this study introduce a novel approach to evaluating cellular behavior in vitro. The concept emerged as a particularly noteworthy outcome. Collectively, these results highlight the substantial potential of AID in advancing the field of cancer cells biology, particularly.

• Klíčová slova
• Biophysics, Cancer cell migration, Digital holographic microscopy, Image processing, Live cell imaging, Non-invasive, Quantitative phase imaging, Staining-free imaging,
• MeSH
• algoritmy MeSH
• časosběrné zobrazování MeSH
• fibrosarkom patologie   MeSH
• holografie metody   MeSH
• lidé MeSH
• mikroskopie MeSH
• nádorové buněčné linie MeSH
• počítačové zpracování obrazu MeSH
• pohyb buněk * MeSH
• programovací jazyk MeSH
• software MeSH
• Check Tag
• lidé MeSH
• Publikační typ
• časopisecké články MeSH

Equivariant neural networks incorporate symmetries into their architecture, achieving higher generalization performance. However, constructing equivariant neural networks typically requires prior knowledge of data types and symmetries, which is difficult to achieve in most tasks. In this paper, we propose LieSD, a method for discovering symmetries via trained neural networks which approximate the input-output mappings of the tasks. It characterizes equivariance and invariance (a special case of equivariance) of continuous groups using Lie algebra and directly solves the Lie algebra space through the inputs, outputs, and gradients of the trained neural network. Then, we extend the method to make it applicable to multi-channel data and tensor data, respectively. We validate the performance of LieSD on tasks with symmetries such as the two-body problem, the moment of inertia matrix prediction, top quark tagging, and rotated MNIST. Compared with the baseline, LieSD can accurately determine the number of Lie algebra bases without the need for expensive group sampling. Furthermore, LieSD can perform well on non-uniform datasets, whereas methods based on GANs fail. Code and data are available at https://github.com/hulx2002/LieSD.

• Klíčová slova
• Equivariant networks, Symmetry discovery,
• MeSH
• algoritmy MeSH
• lidé MeSH
• neuronové sítě * MeSH
• Check Tag
• lidé MeSH
• Publikační typ
• časopisecké články MeSH