NPJ Digit Med. 2025 Feb 21;8(1):118. doi: 10.1038/s41746-025-01507-3.

ABSTRACT

This study forms the basis of a digital twin system of the knee joint, using advanced quantitative MRI (qMRI) and machine learning to advance precision health in osteoarthritis (OA) management and knee replacement (KR) prediction. We combined deep learning-based segmentation of knee joint structures with dimensionality reduction to create an embedded feature space of imaging biomarkers. Through cross-sectional cohort analysis and statistical modeling, we identified specific biomarkers, including variations in cartilage thickness and medial meniscus shape, that are significantly associated with OA incidence and KR outcomes. Integrating these findings into a comprehensive framework represents a considerable step toward personalized knee-joint digital twins, which could enhance therapeutic strategies and inform clinical decision-making in rheumatological care. This versatile and reliable infrastructure has the potential to be extended to broader clinical applications in precision health.

PMID:39984725 | DOI:10.1038/s41746-025-01507-3

Eur J Trauma Emerg Surg. 2025 Feb 21;51(1):119. doi: 10.1007/s00068-025-02793-y.

ABSTRACT

PURPOSE: The application of artificial intelligence (AI) in healthcare has seen widespread implementation, with numerous studies highlighting the development of robust algorithms. However, limited attention has been given to the secure utilization of raw data for medical model training, and its subsequent impact on clinical decision-making and real-world applications. This study aims to assess the feasibility and effectiveness of an advanced diagnostic model that integrates blockchain technology and AI for the identification of tibial plateau fractures (TPFs) in emergency settings.

METHOD: In this study, blockchain technology was utilized to construct a distributed database for trauma orthopedics, images collected from three independent hospitals for model training, testing, and internal validation. Then, a distributed network combining blockchain and deep learning was developed for the detection of TPFs, with model parameters aggregated across multiple nodes to enhance accuracy. The model's performance was comprehensively evaluated using metrics including accuracy, sensitivity, specificity, F1 score, and the area under the receiver operating characteristic curve (AUC). In addition, the performance of the centralized model, the distributed AI model, clinical orthopedic attending physicians, and AI-assisted attending physicians was tested on an external validation dataset.

RESULTS: In the testing set, the accuracy of our distributed model was 0.9603 [95% CI (0.9598, 0.9605)] and the AUC was 0.9911 [95% CI (0.9893, 0.9915)] for TPF detection. In the external validation set, the accuracy reached 0.9636 [95% CI (0.9388, 0.9762)], was slightly higher than that of the centralized YOLOv8n model at 0.9632 [95% CI (0.9387, 0.9755)] (p > 0.05), and exceeded the orthopedic physician at 0.9291 [95% CI (0.9002, 0.9482)] and radiology attending physician at 0.9175 [95% CI (0.8891, 0.9393)], with a statistically significant difference (p < 0.05). Additionally, the centralized model (4.99 ± 0.01 min) had shorter diagnosis times compared to the orthopedic attending physician (25.45 ± 1.92 min) and the radiology attending physician (26.21 ± 1.20 min), with a statistically significant difference (p < 0.05).

CONCLUSION: The model based on the integration of blockchain technology and AI can realize safe, collaborative, and convenient assisted diagnosis of TPF. Through the aggregation of training parameters by decentralized algorithms, it can achieve model construction without data leaving the hospital and may exert clinical application value in the emergency settings.

PMID:39984717 | DOI:10.1007/s00068-025-02793-y

Langenbecks Arch Surg. 2025 Feb 22;410(1):80. doi: 10.1007/s00423-025-03641-8.

ABSTRACT

PURPOSE: Several surgical-skill assessment tools emphasize the importance of efficient tissue-dissection, whose assessment relies on human judgment and is thus subject to bias. Automated assessment may help solve this problem. This study aimed to verify the feasibility of surgical-skill assessment using a deep learning-based recognition model.

METHODS: This retrospective study used multicenter intraoperative videos of laparoscopic colorectal surgery (sigmoidectomy or high anterior resection) for colorectal cancer obtained from 766 cases across Japan. Three groups with different skill levels were distinguished: high-, intermediate-, and low-skill. We developed a model to recognize tissue dissection by the monopolar device using deep learning-based computer-vision technology. Tissue-dissection time per monopolar device appearance time (efficient-dissection time ratio) was extracted as a quantitative parameter describing efficient dissection. We automatically measured the efficient-dissection time ratio using the recognition model of 8 surgical instruments and tissue-dissection on/off classification model. The efficient-dissection time ratio was compared among groups; the feasibility of distinguishing them was explored using the model. The model-calculated parameters were evaluated to determine whether they could differentiate high-, intermediate-, and low-skill groups.

RESULTS: The tissue-dissection recognition model had an overall accuracy of 0.91. There was a moderate correlation (0.542; 95% confidence interval, 0.288-0.724; P < 0.001) between manually and automatically measured efficient-dissection time ratios. Efficient-dissection time ratios by this model were significantly higher in the high-skill than in intermediate-skill (P = 0.0081) and low-skill (P = 0.0249) groups.

CONCLUSION: An automated efficient-dissection assessment model using a monopolar device was constructed with a feasible automated skill-assessment method.

PMID:39984705 | DOI:10.1007/s00423-025-03641-8

Sci Rep. 2025 Feb 21;15(1):6362. doi: 10.1038/s41598-025-90185-y.

ABSTRACT

X-ray inspection is a crucial technique for identifying defects in castings, capable of revealing minute internal flaws such as pores and inclusions. However, traditional methods rely on the subjective judgment of experts, are time-consuming, and prone to errors, which negatively impact the efficiency and accuracy of inspections. Therefore, the development of an automated defect detection model is of significant importance for enhancing the scientific rigor and precision of casting inspections. In this study, we propose a deep learning model specifically designed for detecting small-scale defects in castings. The model employs an end-to-end network architecture and features a loss function based on the Wasserstein distance, which is tailored to optimize the training process for small defect targets, thereby improving detection accuracy. Additionally, we have innovatively developed a dual-layer Encoder-Decoder multi-scale feature extraction architecture, BiSDE, based on the Hadamard product, aimed at enhancing the model's ability to recognize and locate small targets. To evaluate the performance of the proposed model, we conducted a series of experiments, including comparative tests with current state-of-the-art object detection models such as Yolov9, FasterNet, Yolov8, and Detr, as well as ablation studies on the model's components. The results demonstrate that our model achieves at least a 5.3% improvement in Mean Average Precision (MAP) over the existing state-of-the-art models. Furthermore, the inclusion of each component significantly enhanced the overall performance of the model. In conclusion, our research not only validates the effectiveness of the proposed small-scale defect detection model in improving detection precision but also offers broad prospects for the automation and intelligent development of industrial defect inspection.

PMID:39984609 | DOI:10.1038/s41598-025-90185-y

Sci Rep. 2025 Feb 21;15(1):6312. doi: 10.1038/s41598-025-90024-0.

ABSTRACT

Successful translation of artificial intelligence (AI) models into clinical practice, across clinical domains, is frequently hindered by the lack of image quality control. Diagnostic models are often trained on images with no denotation of image quality in the training data; this, in turn, can lead to misclassifications by these models when implemented in the clinical setting. In the case of cervical images, quality classification is a crucial task to ensure accurate detection of precancerous lesions or cancer; this is true for both gynecologic-oncologists' (manual) and diagnostic AI models' (automated) predictions. Factors that impact the quality of a cervical image include but are not limited to blur, poor focus, poor light, noise, obscured view of the cervix due to mucus and/or blood, improper position, and over- and/or under-exposure. Utilizing a multi-level image quality ground truth denoted by providers, we generated an image quality classifier following a multi-stage model selection process that investigated several key design choices on a multi-heterogenous "SEED" dataset of 40,534 images. We subsequently validated the best model on an external dataset ("EXT"), comprising 1,340 images captured using a different device and acquired in different geographies from "SEED". We assessed the relative impact of various axes of data heterogeneity, including device, geography, and ground-truth rater on model performance. Our best performing model achieved an area under the receiver operating characteristics curve (AUROC) of 0.92 (low quality, LQ vs. rest) and 0.93 (high quality, HQ vs. rest), and a minimal total %extreme misclassification (%EM) of 2.8% on the internal validation set. Our model also generalized well externally, achieving corresponding AUROCs of 0.83 and 0.82, and %EM of 3.9% when tested out-of-the-box on the external validation ("EXT") set. Additionally, our model was geography agnostic with no meaningful difference in performance across geographies, did not exhibit catastrophic forgetting upon retraining with new data, and mimicked the overall/average ground truth rater behavior well. Our work represents one of the first efforts at generating and externally validating an image quality classifier across multiple axes of data heterogeneity to aid in visual diagnosis of cervical precancer and cancer. We hope that this will motivate the accompaniment of adequate guardrails for AI-based pipelines to account for image quality and generalizability concerns.

PMID:39984572 | DOI:10.1038/s41598-025-90024-0

Nat Commun. 2025 Feb 21;16(1):1862. doi: 10.1038/s41467-025-57153-6.

ABSTRACT

Genetically engineered commensal bacteria are promising living drugs, however, their therapeutic molecules are frequently confined to their colonization sites. Herein, we report an oral protein delivery technology utilizing an engineered bacterial type zero secretion system (T0SS) via outer membrane vesicles (OMVs). We find that OMVs produced in situ by Escherichia coli Nissle 1917 (EcN) can penetrate the intact gut epithelial barrier to enter the circulation and that epithelial transcytosis involves pinocytosis and dynamin-dependent pathways. EcN is engineered to endogenously load various enzymes into OMVs, and the secreted enzyme-loaded OMVs are able to stably catalyze diverse detoxification reactions against digestive fluid and even enter the circulation. Using hyperuricemic mice and uricase delivery as a demonstration, we demonstrate that the therapeutic efficacy of our engineered EcN with a modified T0SS outperforms that with a direct protein secretion apparatus. The enzyme-loaded OMVs also effectively detoxify human serum samples, highlighting the potential for the clinical treatment of metabolic disorders.

PMID:39984501 | DOI:10.1038/s41467-025-57153-6

Nat Commun. 2025 Feb 21;16(1):1839. doi: 10.1038/s41467-025-56334-7.

ABSTRACT

Shaping embryonic tissues requires spatiotemporal changes in genetic and signaling activity as well as in tissue mechanics. Studies linking specific molecular perturbations to changes in the tissue physical state remain sparse. Here we study how specific genetic perturbations affecting different posterior tissues during zebrafish body axis elongation change their physical state, the resulting large-scale tissue flows, and posterior elongation. Using a custom analysis software to reveal spatiotemporal variations in tissue fluidity, we show that dorsal tissues are most fluid at the posterior end, rigidify anterior of this region, and become more fluid again yet further anteriorly. In the absence of notochord (noto mutants) or when the presomitic mesoderm is substantially reduced (tbx16 mutants), dorsal tissues elongate normally. Perturbations of posterior-directed morphogenetic flows in dorsal tissues (vangl2 mutants) strongly affect the speed of elongation, highlighting the essential role of dorsal cell flows in delivering the necessary material to elongate the axis.

PMID:39984461 | DOI:10.1038/s41467-025-56334-7

Nat Commun. 2025 Feb 21;16(1):1842. doi: 10.1038/s41467-025-56934-3.

ABSTRACT

Antibiotics that target multiple cellular functions are anticipated to be less prone to bacterial resistance. Here we hypothesize that while dual targeting is crucial, it is not sufficient in preventing resistance. Only those antibiotics that simultaneously target membrane integrity and block another cellular pathway display reduced resistance development. To test the hypothesis, we focus on three antibiotic candidates, POL7306, Tridecaptin M152-P3 and SCH79797, all of which fulfill the above criteria. Here we show that resistance evolution against these antibiotics is limited in ESKAPE pathogens, including Escherichia coli, Klebsiella pneumoniae, Acinetobacter baumannii and Pseudomonas aeruginosa, while dual-target topoisomerase antibiotics are prone to resistance. We discover several mechanisms restricting resistance. First, de novo mutations result in only a limited elevation in resistance, including those affecting the molecular targets and efflux pumps. Second, resistance is inaccessible through gene amplification. Third, functional metagenomics reveal that mobile resistance genes are rare in human gut, soil and clinical microbiomes. Finally, we detect rapid eradication of bacterial populations upon toxic exposure to membrane targeting antibiotics. We conclude that resistance mechanisms commonly found in natural bacterial pathogens provide only limited protection to these antibiotics. Our work provides guidelines for the future development of antibiotics.

PMID:39984459 | DOI:10.1038/s41467-025-56934-3

J Proteome Res. 2025 Feb 21. doi: 10.1021/acs.jproteome.4c00677. Online ahead of print.

ABSTRACT

Many factors negatively affect a generalization of the findings in discovery proteomics. They include differentiation between patient cohorts, a variety of experimental conditions, etc. We presented a machine-learning-based workflow for proteomics data analysis, aiming at improving generalizability across multiple data sets. In particular, we customized the decision tree model by introducing a new parameter, min_groups_leaf, which regulates the presence of the samples from each data set inside the model's leaves. Further, we analyzed a trend for the feature importance's curve as a function of the novel parameter for feature selection to a list of proteins with significantly improved generalization. The developed workflow was tested using five proteomic data sets obtained for post-mortem human brain samples of Alzheimer's disease. The data sets consisted of 535 LC-MS/MS acquisition files. The results were obtained for two different pipelines of data processing: (1) MS1-only processing based on DirectMS1 search engine and (2) a standard MS/MS-based one. Using the developed workflow, we found seven proteins with expression patterns that were unique for asymptomatic Alzheimer patients. Two of them, Serotransferrin TRFE and DNA repair nuclease APEX1, may be potentially important for explaining the lack of dementia in patients with the presence of neuritic plaques and neurofibrillary tangles.

PMID:39984290 | DOI:10.1021/acs.jproteome.4c00677

Drug Discov Today. 2025 Feb 19:104315. doi: 10.1016/j.drudis.2025.104315. Online ahead of print.

ABSTRACT

Fucoidan, a polysaccharide from seaweed, holds promise as a drug delivery system and immune modulator; however, its exact mechanism of action remains unclear. As various carbohydrates play key roles in immune responses by binding to carbohydrate-binding proteins like lectins, fucoidan is hypothesized to interact with immune receptors, potentially driving its anticancer activities. However, structural variability, extraction-induced heterogeneity, and weak binding affinities pose challenges to research. Computational tools offer valuable insights into fucoidan-receptor interactions, addressing these challenges and enabling the design of more effective therapies. This review examines fucoidan's therapeutic activities, drug delivery potential, and receptor interactions, emphasizing computational approaches to advance immune modulation and anticancer applications using carbohydrate polymers.

PMID:39984116 | DOI:10.1016/j.drudis.2025.104315

Am J Hum Genet. 2025 Feb 18:S0002-9297(25)00044-8. doi: 10.1016/j.ajhg.2025.01.024. Online ahead of print.

ABSTRACT

Congenital heart disease (CHD) is the most common congenital anomaly and a leading cause of infant morbidity and mortality. Despite extensive exploration of the monogenic causes of CHD over the last decades, ∼55% of cases still lack a molecular diagnosis. Investigating digenic interactions, the simplest form of oligogenic interactions, using high-throughput sequencing data can elucidate additional genetic factors contributing to the disease. Here, we conducted a comprehensive analysis of digenic interactions in CHD by utilizing a large CHD trio exome sequencing cohort, comprising 3,910 CHD and 3,644 control trios. We extracted pairs of presumably deleterious rare variants observed in CHD-affected and unaffected children but not in a single parent. Burden testing of gene pairs derived from these variant pairs revealed 29 nominally significant gene pairs. These gene pairs showed a significant enrichment for known CHD genes (p < 1.0 × 10-4) and exhibited a shorter average biological distance to known CHD genes than expected by chance (p = 3.0 × 10-4). Utilizing three complementary biological relatedness approaches including network analyses, biological distance calculations, and candidate gene prioritization methods, we prioritized 10 final gene pairs that are likely to underlie CHD. Analysis of bulk RNA-sequencing data showed that these genes are highly expressed in the developing embryonic heart (p < 1 × 10-4). In conclusion, our findings suggest the potential role of digenic interactions in CHD pathogenesis and provide insights into unresolved molecular diagnoses. We suggest that the application of the digenic approach to additional disease cohorts will significantly enhance genetic discovery rates.

PMID:39983722 | DOI:10.1016/j.ajhg.2025.01.024

Dev Cell. 2025 Feb 18:S1534-5807(25)00059-0. doi: 10.1016/j.devcel.2025.01.018. Online ahead of print.

ABSTRACT

The neural crest is a highly plastic stem cell population that represents an exception to the germ layer theory. Despite being of ectodermal origin, cranial neural crest cells can differentiate into skeletal derivatives typically formed by mesoderm. Here, we report that SMAD2/3-mediated transforming growth factor β (TGF-β) signaling enhances neural crest developmental potential in the chicken embryo. Our results show that TGF-β signaling modulates neural crest axial identity and directly controls the gene circuits that support skeletal differentiation. Cooperation between TGF-β and low levels of WNT signaling in the embryonic head activates cranial-specific cis-regulatory elements. Activation of TGF-β signaling reprogrammed trunk neural crest cells into adopting an anterior identity and led to the development of an improved protocol for the generation of human cranial neural crest cells. Our findings indicate TGF-β signaling is required for the specification of cranial neural crest cells, endowing them with the potential to give rise to the craniofacial skeleton.

PMID:39983721 | DOI:10.1016/j.devcel.2025.01.018

Cell. 2025 Feb 20;188(4):958-977.e23. doi: 10.1016/j.cell.2024.11.034.

ABSTRACT

Membrane contact sites (MCSs) are fundamental for intracellular communication, but their role in intercellular communication remains unexplored. We show that in plants, plasmodesmata communication bridges function as atypical endoplasmic reticulum (ER)-plasma membrane (PM) tubular MCSs, operating at cell-cell interfaces. Similar to other MCSs, ER-PM apposition is controlled by a protein-lipid tethering complex, but uniquely, this serves intercellular communication. Combining high-resolution microscopy, molecular dynamics, and pharmacological and genetic approaches, we show that cell-cell trafficking is modulated through the combined action of multiple C2 domains transmembrane domain proteins (MCTPs) 3, 4, and 6 ER-PM tethers and phosphatidylinositol-4-phosphate (PI4P) lipid. Graded PI4P amounts regulate MCTP docking to the PM, their plasmodesmata localization, and cell-cell permeability. SAC7, an ER-localized PI4P-phosphatase, regulates MCTP4 accumulation at plasmodesmata and modulates cell-cell trafficking capacity in a cell-type-specific manner. Our findings expand MCS functions in information transmission from intracellular to intercellular cellular activities.

PMID:39983675 | DOI:10.1016/j.cell.2024.11.034

J Plant Physiol. 2025 Feb 15;306:154453. doi: 10.1016/j.jplph.2025.154453. Online ahead of print.

ABSTRACT

An interesting aspect that links the geomagnetic field (GMF) to the evolution of life lies in how plants respond to the reduction of the GMF, also known as hypomagnetic field (HMF). In this work, tomato plants (Solanum lycopersicum cv Microtom) were exposed either to GMF or HMF and were studied during the development of leaves and fruit set. Changes of expression of genes encoding for primary and secondary metabolites, including Reactive Oxygen Species (ROS), proteins, fatty acids, polyphenols, chlorophylls, carotenoids and phytohormones were assessed by qRT-PCR, while the corresponding metabolite levels were quantified by GC-MS and HPLC-MS. Two tomato homologs of the fruit fly magnetoreceptor MagR, Isca-like 1 and erpA 2, were modulated by HMF, as were numerous tomato genes under investigation. In tomato leaves, positive correlations were observed with most of the genes associated with phytohormones production, ROS scavenging and production, and lipid metabolism, whereas an almost reversed trend was found in flowers and fruits. Interestingly, downregulation of Isca-like 1 and erpA 2 was found to correlate with an upregulation of most unripe fruit genes. Exposure to HMF reduced chlorophyll and carotenoid content, decreased photosynthetic efficiency and increased non-photochemical quenching. Auxins, gibberellins, cytokinins, abscisic acid, jasmonic acid and salicylic acid content and the expression of genes related to their metabolism correlated with tomato ISCA modulation. The results here reported suggest that Isca-like 1 and erpA 2 might be important players in tomato magnetoreception.

PMID:39983659 | DOI:10.1016/j.jplph.2025.154453

Sci Rep. 2025 Feb 21;15(1):6310. doi: 10.1038/s41598-025-91102-z.

ABSTRACT

Immune related adverse events (irAEs) occur due to the inflammatory side effects of immune check point inhibitors (ICIs) and irAEs have been associated with improved efficacy in advanced non-small lung cancer (NSCLC) patients. Corticosteroids can reduce the efficacy of ICIs due to their immunosuppressive effects. In this study, we aimed to show the effects of the development of irAEs and the use of ≥ 10 mg prednisone and equivalent steroids on treatment response. We analyzed the outcomes of patients with NSCLC treated with ICIs as monotherapy or ICIs in combination with chemotherapy (ChT) at a single academic center based on the presence of irAEs and the use of corticosteroids. A landmark analysis was performed due to the time-dependent nature of irAEs. 90 patients were included in the study. irAEs were seen in a total of 45 (50%) patients. In the landmark analysis, the median overall survival (OS) was 52.1 months in those who developed irAEs and 14.4 months in those who did not (HR 2.71, 95% CI (1.55-4.73), p < 0.001), and the median progression-free survival (PFS) was also higher in the patients with irAEs (25.9 vs. 8.4 months, HR 2.54, 95% CI (1.52-4.25), p < 0.001). The objective response rate (ORR) was significantly higher in patients experiencing irAEs than without irAEs: 60% versus 33.3%, respectively (p = 0.011). The number of patients using steroids was 22 (24%), while 68 patients (76%) were not using steroids. There was no significant difference in mOS: 26.5 versus 28.7 months (HR 1.14, 95% CI (0.63-2.08), p = 0.652) and mPFS: 16.9 versus 13.5 months (HR 0.99, 95% CI (0.57-1.74), p = 0.997) between patients who used steroids and those who did not. ICIs efficacy is higher in patients who developed irAEs. In our analyses, the grade of irAEs or the number of irAEs occurring in the individual had no effect on mOS and mPFS. In our patient group, steroid use was mostly related with irAEs, and we did not detect any negative effects of corticosteroid use on PFS and OS.

PMID:39984593 | DOI:10.1038/s41598-025-91102-z

Pharmacol Res Perspect. 2025 Apr;13(2):e70063. doi: 10.1002/prp2.70063.

ABSTRACT

Interstitial lung disease (ILD) is a clinically relevant adverse event associated with biologic agent use. However, the current incidence of ILD remains unclear as large-scale risk assessments of biologic agents have not been conducted. The aim of this study was to clarify the association between biologic agent use and ILD development in clinical practice by detecting adverse event signals using a spontaneous adverse drug reaction database. The VigiBase database is used for spontaneous adverse event reporting. The analysis focused on nine biologics used to treat psoriasis, rheumatoid arthritis, and Crohn's disease. The safety of each biologic agent was evaluated using the information component signal detection method. There were 32,520,983 reports in VigiBase, of which 68,489 (0.21%) were for ILD. Signals were mainly detected for tumor necrosis factor-α inhibitors when the information component for ILD caused by biologic agents was calculated. Comorbidity analysis in patients who developed ILD and analysis of the time from the start of treatment with each drug to ILD onset showed differences for each biologic agent. ILD is a serious adverse effect of biologic agents, and there are several cases in which a causal relationship with ILD development cannot be ruled out. The occurrence of interstitial ILD should be noted when using biologics, particularly TNF-α inhibitors.

PMID:39984304 | DOI:10.1002/prp2.70063

Am J Physiol Lung Cell Mol Physiol. 2025 Feb 21. doi: 10.1152/ajplung.00293.2024. Online ahead of print.

ABSTRACT

Bronchial epithelial cells derived from the tracheo-bronchial regions of human airways (HBECs) provide a valuable in vitro model for studying pathological mechanisms and evaluating therapeutics. This cell population comprises a mixed population of basal cells (BCs), the predominant stem cell in airways capable of both self-renewal and functional differentiation. Despite their potential for regenerative medicine, BCs exhibit significant phenotypic variability in culture. To investigate how culture conditions influence BC phenotype and function, we expanded three independent BC isolates in three media: airway epithelial cell growth medium (AECGM), dual-SMAD inhibitor (DSI)-enriched AECGM, and Pneumacult Ex plus (PEx+). Analysis through RNA sequencing, immune assays and impedance measurements revealed that PEx+ media significantly drove cell proliferation and a broad pro-inflammatory phenotype in BCs. In contrast, BCs expanded in AECGM, displayed increased expression of structural and extracellular matrix components at higher passage. AECGM increased expression of some cytokines at high passage, while DSI suppressed inflammation implicating the involvement TGF-β in BC inflammatory processes. Differentiation capacity of BCs declined with time in culture irrespective of expansion media. This was associated with an increase in PLUNC expressing secretory cells in AECGM and PEx+ media consistent with the known immune modulatory role of PLUNC in the airways. These findings highlight the profound impact of media conditions on inflammatory niche established by, and function of, in vitro expanded BCs. The broad pro-inflammatory phenotype driven by PEx+ media, in particular, should be considered in the development of cell-based models for airway diseases and therapeutic application.

PMID:39982813 | DOI:10.1152/ajplung.00293.2024

Biomed Phys Eng Express. 2025 Feb 21. doi: 10.1088/2057-1976/adb8f1. Online ahead of print.

ABSTRACT

OBJECTIVE: To determine the minimum detector width along the fan-angle direction in offset-detector cone-beam breast CT for multiple advanced reconstruction algorithms and to investigate the effect on radiation dose, scatter, and image quality.

APPROACH: Complete sinograms (m × n = 1024 × 768 pixels) of 30 clinical breast CT datasets previously acquired on a clinical-prototype cone-beam breast CT system were reconstructed using Feldkamp-Davis-Kress (FDK) algorithm and served as the reference. Complete sinograms were retrospectively truncated to varying widths to understand the limits of four image reconstruction algorithms - FDK with redundancy weighting (FDK-W), compressed-sensing based FRIST, fully-supervised MS-RDN, and self-supervised AFN. Upon determining the truncation limits, numerical phantoms generated by segmenting the reference reconstructions into skin, adipose, and fibroglandular tissues were used to determine the radiation dose and scatter-to-primary ratio (SPR) using Monte Carlo simulations.

MAIN RESULTS: FDK-W, FRIST, and MS-RDN showed artifacts when m < 596, whereas AFN reconstructed images without artifacts for m>=536. Reducing the detector width reduced signal-difference to noise ratio (SDNR) for FDK-W, whereas FRIST, MS-RDN and AFN maintained or improved SDNR. Reference reconstruction and AFN with m=536 had similar quantitative measures of image quality.

SIGNIFICANCE: For the 30 cases, AFN with m=536 reduced the radiation dose and SPR by 37.85% and 33.46%, respectively, compared to the reference. Qualitative and quantitative image quality indicate the feasibility of AFN for offset-detector cone-beam breast CT. Radiation dose and SPR were simultaneously reduced with a 536 ×768 detector and when used in conjunction with AFN algorithm had similar image quality as the reference reconstruction.

PMID:39983239 | DOI:10.1088/2057-1976/adb8f1

J Neural Eng. 2025 Feb 21. doi: 10.1088/1741-2552/adb90c. Online ahead of print.

ABSTRACT

OBJECTIVE: Humans spend a significant portion of their lives in sleep (an essential driver of body metabolism). Moreover, as sleep deprivation could cause various health complications, it is crucial to develop an automatic sleep stage detection model to facilitate the tedious manual labeling process. Notably, recently proposed sleep staging algorithms lack model explainability and still require performance improvement.

APPROACH: We implemented multiscale neurophysiology-mimicking kernels to capture sleep-related electroencephalogram (EEG) activities at varying frequencies and temporal lengths; the implemented model was named "Multiscale Temporal Convolutional Neural Network (MTCNN)." Further, we evaluated its performance using an open-source dataset (Sleep-EDF Database Expanded comprising 153 days of polysomnogram data).

MAIN RESULTS: By investigating the learned kernel weights, we observed that MTCNN detected the EEG activities specific to each sleep stage, such as the frequencies, K-complexes, and sawtooth waves. Furthermore, regarding the characterization of these neurophysiologically significant features, MTCNN demonstrated an overall accuracy (OAcc) of 91.12% and a Cohen kappa coefficient of 0.86 in the cross-subject paradigm. Notably, it demonstrated an OAcc of 88.24% and a Cohen kappa coefficient of 0.80 in the leave-few-days-out analysis. Our MTCNN model also outperformed the existing deep learning models in sleep stage classification even when it was trained with only 16% of the total EEG data, achieving an OAcc of 85.62% and a Cohen kappa coefficient of 0.75 on the remaining 84% of testing data.

SIGNIFICANCE: The proposed MTCNN enables model explainability and it can be trained with lesser amount of data, which is beneficial to its application in the real-world because large amounts of training data are not often and readily available.

PMID:39983236 | DOI:10.1088/1741-2552/adb90c

J Chem Inf Model. 2025 Feb 21. doi: 10.1021/acs.jcim.4c01371. Online ahead of print.

ABSTRACT

The importance of drug toxicity assessment lies in ensuring the safety and efficacy of the pharmaceutical compounds. Predicting toxicity is crucial in drug development and risk assessment. This study compares the performance of GPT-4 and GPT-4o with traditional deep-learning and machine-learning models, WeaveGNN, MorganFP-MLP, SVC, and KNN, in predicting molecular toxicity, focusing on bone, neuro, and reproductive toxicity. The results indicate that GPT-4 is comparable to deep-learning and machine-learning models in certain areas. We utilized GPT-4 combined with molecular docking techniques to study the cardiotoxicity of three specific targets, examining traditional Chinese medicinal materials listed as both food and medicine. This approach aimed to explore the potential cardiotoxicity and mechanisms of action. The study found that components in Black Sesame, Ginger, Perilla, Sichuan Pagoda Tree Fruit, Galangal, Turmeric, Licorice, Chinese Yam, Amla, and Nutmeg exhibit toxic effects on cardiac target Cav1.2. The docking results indicated significant binding affinities, supporting the hypothesis of potential cardiotoxic effects.This research highlights the potential of ChatGPT in predicting molecular properties and its significance in medicinal chemistry, demonstrating its facilitation of a new research paradigm: with a data set, high-accuracy learning models can be generated without requiring computational knowledge or coding skills, making it accessible and easy to use.

PMID:39982968 | DOI:10.1021/acs.jcim.4c01371