Biomedicines. 2025 Mar 29;13(4):822. doi: 10.3390/biomedicines13040822.

ABSTRACT

Background: Optimizing sedative techniques for drug-induced sleep endoscopy (DISE) enhances accuracy and reproducibility in tailoring treatment for obstructive sleep apnea (OSA). The Schnider and Eleveld pharmacokinetic-pharmacodynamic (PK-PD) models, which predict propofol concentration in effect-site compartment based on patient-specific parameters, were utilized to guide intravenous sedation in this study. We compared the effectiveness of propofol sedation guided by the novel general-purpose Eleveld model versus the Schnider model using target-controlled infusion (TCI) systems. Methods: We investigated twenty-five adult OSA patients, randomized into two groups: the Schnider model group (n = 12) and the Eleveld model group (n = 13). DISE was conducted following standardized protocols, targeting effect-site concentration TCI mode. Data concerning sedation levels, effect-site concentration of propofol, procedural timing, propofol dosages, respiratory and cardiovascular parameters, and any procedural incidents were collected. Results: DISE was performed successfully in all enrolled patients from both groups. A significant difference was observed in the effect-site concentration of propofol (CeP) at the moment of endoscopy between the Eleveld and Schnider groups (2.1 ± 0.4 µg/mL vs. 3.3 ± 0.7 µg/mL, respectively; p < 0.001). The E group also demonstrated a shorter time to attain the optimal sedation plane compared to the S group (6.1 ± 1.7 vs. 9.8 ± 2.2 min, respectively; p < 0.001) and a reduced total procedural time (11.2 ± 1.4 vs. 15.0 ± 2.1 min, respectively; p < 0.001). The incidence of adverse events was comparable between groups. Conclusions: The Eleveld model demonstrated a shorter time to achieve the optimal sedation plane, a shorter total procedural time, and a significant difference in effect-site concentration at the time of endoscopy compared to the Schnider model. The incidence of adverse events was comparable between the two groups, suggesting that the Eleveld model may offer improved efficiency without compromising safety during DISE.

PMID:40299425 | DOI:10.3390/biomedicines13040822

medRxiv [Preprint]. 2025 Apr 8:2025.04.07.25325038. doi: 10.1101/2025.04.07.25325038.

ABSTRACT

The urgent need for safe and effective therapies for Alzheimer's disease (AD) has spurred a growing interest in repurposing existing drugs to treat or prevent AD. In this study, we combined multi-omics and clinical data to investigate possible repurposing opportunities for AD. We performed transcriptome-wide association studies (TWAS) to construct gene expression signatures of AD from publicly available GWAS summary statistics, using both transcriptome prediction models for 49 tissues from the Genotype-Tissue Expression (GTEx) project and microglia-specific models trained on eQTL data from the Microglia Genomic Atlas (MiGA). We then identified compounds capable of reversing the AD-associated changes in gene expression observed in these signatures by querying the Connectivity Map (CMap) drug perturbation database. Out of >2,000 small-molecule compounds in CMap, aspirin emerged as the most promising AD repurposing candidate. To investigate the longitudinal effects of aspirin use on AD, we collected drug exposure and AD coded diagnoses from three independent sources of real-world data: electronic health records (EHRs) from Vanderbilt University Medical Center (VUMC) and the National Institutes of Health All of Us Research Program, along with national healthcare claims from the MarketScan Research Databases. In meta-analysis of EHR data from VUMC and All of Us , we found that aspirin use before age 65 was associated with decreased risk of incident AD (hazard ratio=0.76, 95% confidence interval [CI]: 0.64-0.89, P =0.001). Consistent with the findings utilizing EHR data, analysis of claims data from MarketScan revealed significantly lower odds of aspirin exposure among AD cases compared to matched controls (odds ratio=0.32, 95% CI: 0.28-0.38, P <0.001). Our results demonstrate the value of integrating genetic and clinical data for drug repurposing studies and highlight aspirin as a promising repurposing candidate for AD, warranting further investigation in clinical trials.

PMID:40297415 | PMC:PMC12036415 | DOI:10.1101/2025.04.07.25325038

Clin Transl Sci. 2025 May;18(5):e70225. doi: 10.1111/cts.70225.

ABSTRACT

Simvastatin, an HMG-CoA reductase inhibitor, is widely used for hypercholesterolemia but may cause myotoxicity linked to its plasma concentration. Pharmacokinetic gene polymorphisms influence inter-individual variability in simvastatin exposure. This study investigated the effects of pharmacokinetic gene polymorphisms on steady-state simvastatin plasma levels in Thai patients. Eighty-nine Thai patients with dyslipidemia or coronary artery disease on simvastatin treatment for at least 2 weeks without dose adjustment were recruited from King Chulalongkorn Memorial Hospital. Simvastatin lactone and acid concentrations were measured 12 h post-dose using UHPLC-MS/MS. Pharmacokinetic gene polymorphisms, including ABCB1, ABCC2, ABCG2, SLCO1B1, SLCO1B3, CYP3A4, and CYP3A5, were genotyped by MassARRAY System. The results showed that patients with the SLCO1B1 c.521TC+CC genotype had significantly higher simvastatin acid levels than those with c.521TT (0.53 vs. 0.19 ng/mL, p = 0.03). Similarly, the SLCO1B1*1b/*15 genotype was associated with higher simvastatin acid levels than SLCO1B1*1a/*1a (0.58 vs. 0.16 ng/mL, p < 0.001). These findings suggest that SLCO1B1 c.521T>C, alone or with c.388A>G (SLCO1B1*1b/*15), reduces OATP1B1 function, leading to elevated simvastatin acid levels and increased myotoxicity risk. This study confirms the association of SLCO1B1 rs4149056 (c.521T>C) with higher simvastatin plasma levels in Thai patients. The study highlights the potential role of SLCO1B1 genotyping, particularly rs4149056 (c.521T>C) and rs2306283 (c.388A>G), in guiding statin therapy for Thai patients, which could help optimize treatment and reduce adverse effects such as statin-induced myotoxicity.

PMID:40297930 | DOI:10.1111/cts.70225

Diabetes Metab Syndr Obes. 2025 Apr 23;18:1255-1262. doi: 10.2147/DMSO.S500336. eCollection 2025.

ABSTRACT

BACKGROUND/OBJECTIVES: This subgroup analysis of a randomised, open-label, two-period crossover trial in Aotearoa New Zealand (February 2019 to March 2020) assessed whether the glucose-lowering effects of vildagliptin, vs pioglitazone varied by the CREBRF (p.Arg457Gln) rs373863828 genotype.

METHODS: Adults with type 2 diabetes and HbA1c > 58 mmol/mol (>7.5%) received either pioglitazone (30 mg) or vildagliptin (50 mg) for 16 weeks, then switched medications for another 16 weeks. Differences in HbA1c between treatments (pioglitazone vs vildagliptin) were tested for an interaction with CREBRF rs373863828 A-allele carrier status and controlling for baseline HbA1c using linear mixed models. Secondary endpoints included weight, systolic blood pressure, and diabetes treatment satisfaction.

RESULTS: Participants with the AA/AG genotype had a higher baseline weight than those with the GG genotype (121.4 kg vs 106.6 kg, respectively; p<0.01). No significant difference in achieved HbA1c was found based on A-allele carrier status (0.43 mmol/mol; 95% CI -4.83, 5.69; p=0.87). Among Māori and Pacific participants with the A-allele, a smaller weight difference was observed after pioglitazone vs vildagliptin compared to those with the GG genotype (interaction effect -1.66 kg; 95% CI -3.27, -0.05; p=0.04).

CONCLUSION: CREBRF rs373863828 A-allele carriers show a similar HbA1c-lowering response to pioglitazone vs vildagliptin compared to non-carriers but exhibit less weight gain with pioglitazone, despite having significantly higher baseline weights.

PMID:40297769 | PMC:PMC12035406 | DOI:10.2147/DMSO.S500336

Front Pharmacol. 2025 Apr 14;16:1524623. doi: 10.3389/fphar.2025.1524623. eCollection 2025.

ABSTRACT

INTRODUCTION: Thyroid cancer, a prevalent endocrine malignancy, has an age-standardized incidence rate of 9.1 per 100,000 people and a mortality rate of 0.44 per 100,000 as of 2024. Despite significant advances in precision oncology driven by large-scale international consortia, gaps persist in understanding the genomic landscape of thyroid cancer and its impact on therapeutic efficacy across diverse populations.

METHODS: To address this gap, we performed comprehensive data mining and in silico analyses to identify pathogenic variants in thyroid cancer driver genes, calculate allele frequencies, and assess deleteriousness scores across global populations, including African, Amish, Ashkenazi Jewish, East and South Asian, Finnish and non-Finnish European, Latino, and Middle Eastern groups. Additionally, pharmacogenomic profiling, in silico drug prescription, and clinical trial data were analyzed to prioritize targeted therapeutic strategies.

RESULTS: Our analysis examined 56,622 variants in 40 thyroid cancer-driver genes across 76,156 human genomes, identifying 5,001 known and predicted oncogenic variants. Enrichment analysis revealed critical pathways such as MAPK, PI3K-AKT-mTOR, and p53 signaling, underscoring their roles in thyroid cancer pathogenesis. High-throughput validation strategies confirmed actionable genomic alterations in RET, BRAF, NRAS, KRAS, and EPHA7. Ligandability assessments identified these proteins as promising therapeutic targets. Furthermore, our findings highlight the clinical potential of targeted drug inhibitors, including vandetanib, dabrafenib, and selumetinib, for improving treatment outcomes.

DISCUSSION: This study underscores the significance of integrating genomic insights with pharmacogenomic strategies to address disparities in thyroid cancer treatment. The identification of population-specific oncogenic variants and actionable therapeutic targets provides a foundation for advancing precision oncology. Future efforts should focus on including underrepresented populations, developing population-specific prevention strategies, and fostering global collaboration to ensure equitable access to pharmacogenomic testing and innovative therapies. These initiatives have the potential to transform thyroid cancer care and align with the broader goals of personalized medicine.

PMID:40297138 | PMC:PMC12034932 | DOI:10.3389/fphar.2025.1524623

Acta Biochim Biophys Sin (Shanghai). 2025 Apr 28. doi: 10.3724/abbs.2025013. Online ahead of print.

ABSTRACT

Platinum-based chemotherapy remains the mainstay for non-small cell lung cancer (NSCLC), but it frequently causes dose-limiting myelosuppression, with significant individual variability in susceptibility. However, the genetic basis of myelosuppression side effects remains elusive, greatly hindering personalized therapeutic approaches. In this study, we perform a comprehensive genome-wide association analysis on 491 NSCLC patients receiving platinum-based chemotherapy, examining 4,690,998 single-nucleotide polymorphisms (SNPs) to identify relevant genetic variants. LDBlockShow, FUMA, and MAGMA are utilized to explore linkage disequilibrium, expression quantitative trait loci (eQTLs), chromatin interaction, and conduct gene-based and gene set-based analysis of candidate SNPs. The GWAS results reveal that rs6856089 and its linked SNPs are significantly associated with platinum-based chemotherapy-induced myelosuppression. Specifically, patients with the A allele of rs6856089 have a significantly lower risk of myelosuppression (odds ratio (OR) = 0.1300, P = 7.59 × 10 -8). Furthermore, gene-based analysis reveals that EMCN ( P = 2.47 × 10 -5), which encodes endomucin, a marker for hematopoietic stem cells, might mediate myelosuppression. This study provides a scientific basis for the individual differences in platinum-based chemotherapy-induced myelosuppression.

PMID:40296719 | DOI:10.3724/abbs.2025013

Vavilovskii Zhurnal Genet Selektsii. 2025 Apr;29(2):279-289. doi: 10.18699/vjgb-25-31.

ABSTRACT

Cystic fibrosis (CF) is a disease with a broad clinical and genetic spectrum of manifestations, significantly impacting the quality and duration of life of patients. At present, a diagnosis of CF enables the disease to be identified at the earliest stages of its development. The accelerated advancement of scientific knowledge and contemporary research techniques has transformed the methodology employed in the treatment of CF, encompassing a spectrum of approaches from symptomatic management to pathogenetic therapies. Pathogenetic therapy represents an approach to treatment that aims to identify methods of restoring the function of the CFTR gene. The objective of this review was to analyse and summarize the available scientific data on the pathogenetic therapy of CF. This paper considers various approaches to the pathogenetic therapy of CF that are based on the use of targeted drugs known as CFTR modulators. The article presents studies employing gene therapy techniques for CF, which are based on the targeted delivery of a normal copy of the CFTR gene cDNA to the respiratory tract via viral or non-viral vectors. Some studies have demonstrated the efficacy of RNA therapeutic interventions in restoring splicing, promoting the production of mature RNA, and increasing the functional expression of the CFTR protein. The review also analyzes literature data that consider methods of etiotropic therapy for CF, which consists of targeted correction of the CFTR gene using artificial restriction enzymes, the CRISPR/Cas9 system and a complex of peptide-nucleic acids. In a prospective plan, the use of cell therapy methods in the treatment of lung damage in CF is considered.

PMID:40297296 | PMC:PMC12036567 | DOI:10.18699/vjgb-25-31

J Pathol Inform. 2025 Mar 27;17:100438. doi: 10.1016/j.jpi.2025.100438. eCollection 2025 Apr.

ABSTRACT

Immune cell differentials are most commonly performed manually or with the use of automated cell sorting devices. However, manual review by research personnel can be both subjective and time consuming, and cell sorting approaches consume samples and demand additional reagents to perform the differential. We have created an artificial intelligence (AI) image processing pipeline using the Biodock.ai platform to classify immune cell types from Giemsa-stained cytospins of mouse bronchoalveolar lavage fluid. Through multiple rounds of training and refinement, we have created a tool that is as accurate as manual review of slide images while removing the subjectivity and making the process mostly hands off, saving researcher time for other tasks and improving core turnaround for experiments. This AI-based image processing is directly compatible with current workflows utilizing stained slides, in contrast to a change to a flow cytometry-based approach, which requires both specialized equipment, reagents, and expertise.

PMID:40297061 | PMC:PMC12036075 | DOI:10.1016/j.jpi.2025.100438

Res Sq [Preprint]. 2025 Apr 7:rs.3.rs-6149503. doi: 10.21203/rs.3.rs-6149503/v1.

ABSTRACT

Recently, topological deep learning (TDL), which integrates algebraic topology with deep neural networks, has achieved tremendous success in processing point-cloud data, emerging as a promising paradigm in data science. However, TDL has not been developed for data on differentiable manifolds, including images, due to the challenges posed by differential topology. We address this challenge by introducing manifold topological deep learning (MTDL) for the first time. To highlight the power of Hodge theory rooted in differential topology, we consider a simple convolutional neural network (CNN) in MTDL. In this novel framework, original images are represented as smooth manifolds with vector fields that are decomposed into three orthogonal components based on Hodge theory. These components are then concatenated to form an input image for the CNN architecture. The performance of MTDL is evaluated using the MedMNIST v2 benchmark database, which comprises 717,287 biomedical images from eleven 2D and six 3D datasets. MTDL significantly outperforms other competing methods, extending TDL to a wide range of data on smooth manifolds.

PMID:40297704 | PMC:PMC12036455 | DOI:10.21203/rs.3.rs-6149503/v1

Head Neck Tumor Segm MR Guid Appl (2024). 2025;15273:99-111. doi: 10.1007/978-3-031-83274-1_7. Epub 2025 Mar 3.

ABSTRACT

Accurate segmentation of gross tumor volume (GTV) is essential for effective MRI-guided adaptive radiotherapy (MRgART) in head and neck cancer. However, manual segmentation of the GTV over the course of therapy is time-consuming and prone to interobserver variability. Deep learning (DL) has the potential to overcome these challenges by automatically delineating GTVs. In this study, our team, UW LAIR, tackled the challenges of both pre-radiotherapy (pre-RT) (Task 1) and mid-radiotherapy (mid-RT) (Task 2) tumor volume segmentation. To this end, we developed a series of DL models for longitudinal GTV segmentation. The backbone of our models for both tasks was SegResNet with deep supervision. For Task 1, we trained the model using a combined dataset of pre-RT and mid-RT MRI data, which resulted in the improved aggregated Dice similarity coefficient (DSCagg) on a hold-out internal testing set compared to models trained solely on pre-RT MRI data. In Task 2, we introduced mask-aware attention modules, enabling pre-RT GTV masks to influence intermediate features learned from mid-RT data. This attention-based approach yielded slight improvements over the baseline method, which concatenated mid-RT MRI with pre-RT GTV masks as input. In the final testing phase, the ensemble of 10 pre-RT segmentation models achieved an average DSCagg of 0.794, with 0.745 for primary GTV (GTVp) and 0.844 for metastatic lymph nodes (GTVn) in Task 1. For Task 2, the ensemble of 10 mid-RT segmentation models attained an average DSCagg of 0.733, with 0.607 for GTVp and 0.859 for GTVn, leading us to achieve 1st place. In summary, we presented a collection of DL models that could facilitate GTV segmentation in MRgART, offering the potential to streamline radiation oncology workflows.

PMID:40297614 | PMC:PMC12036643 | DOI:10.1007/978-3-031-83274-1_7

medRxiv [Preprint]. 2025 Apr 11:2025.04.08.25325262. doi: 10.1101/2025.04.08.25325262.

ABSTRACT

Late-onset Tay-Sachs (LOTS) disease and late-onset Sandhoff disease (LOSD) have long been considered indistinguishable due to similar clinical presentations and shared biochemical deficits. However, recent magnetic resonance imaging (MRI) studies have shown distinct cerebellar atrophy associated with LOTS. In this study, we furthered this investigation to determine if the cerebellar atrophy is globally uniform or preferentially targets certain cerebellar regions. We utilized DeepCERES , a deep learning cerebellar specific segmentation and cortical thickness pipeline to analyze differences between LOTS (n=20), LOSD (n=5), and neurotypical controls (n=1038). LOTS had smaller volumes of the whole cerebellum as well as cerebellar lobules IV, V, VI, VIIB, VIIIA, VIIIB, IX, and both Crus I and II compared to both LOSD and neurotypical controls. LOTS patients also had smaller cortical thickness of cerebellar lobules V, VI, VIIB, VIIIA, VIIIB, and both Crus I and II compared to both LOSD and neurotypical controls. Cerebellar functional and lesion localization studies have implicated lobules V and VI in speech articulation and execution while lobules VI, Crus I, VIIA, among others, have been implicated in a variety of behaviors and neuropsychiatric symptoms. Our observations provide a possible anatomical substrate to the higher prevalence of dysarthria and psychosis in our LOTS but not LOSD patients. Future studies are needed for direct comparisons considering phenotypic aspects such as age of symptom onset, presence and severity of dysarthria and ataxia, full characterization of neuropsychiatric profiles, molecular pathology and biochemical differences to fully understand the dichotomy observed in these two diseases.

PMID:40297453 | PMC:PMC12036421 | DOI:10.1101/2025.04.08.25325262

medRxiv [Preprint]. 2025 Apr 7:2025.04.06.25325337. doi: 10.1101/2025.04.06.25325337.

ABSTRACT

Alzheimer's disease (AD) and AD-related dementias (ADRD) exhibit heterogeneous progression rates, with rapid progression (RP) posing significant challenges for timely intervention and treatment. The increasingly available patient-centered electronic health records (EHRs) have made it possible to develop advanced machine learning models for risk prediction of disease progression by leveraging comprehensive clinical, demographic, and laboratory data. In this study, we propose AutoRADP, an interpretable autoencoder-based framework that predicts rapid AD/ADRD progression using both structured and unstructured EHR data from UFHealth. AutoRADP incorporates a rule-based natural language processing method to extract critical cognitive assessments from clinical notes, combined with feature selection techniques to identify essential structured EHR features. To address the data imbalance issue, we implement a hybrid sampling strategy that combines similarity-based and clustering-based upsampling. Additionally, by utilizing SHapley Additive exPlanations (SHAP) values, we provide interpretable predictions, shedding light on the key factors driving the rapid progression of AD/ADRD. We demonstrate that AutoRADP outperforms existing methods, highlighting the potential of our framework to advance precision medicine by enabling accurate and interpretable predictions of rapid AD/ADRD progression, and thereby supporting improved clinical decision-making and personalized interventions.

PMID:40297450 | PMC:PMC12036374 | DOI:10.1101/2025.04.06.25325337

medRxiv [Preprint]. 2025 Apr 8:2025.04.07.25325386. doi: 10.1101/2025.04.07.25325386.

ABSTRACT

Alzheimer's disease (AD), particularly late-onset AD, stands as the most prevalent neurodegenerative disorder globally. Owing to its substantial heritability, genetic studies have emerged as indispensable for elucidating genes and biological pathways driving AD onset and progression. However, genetic and molecular mechanisms underlying AD remain poorly defined, largely due to the pronounced heterogeneity of AD and the intricate interactions among AD genetic factors. Notably, approximately 90% of AD-associated genetic variants reside in intronic and intergenic regions, yet their functional significance has remained largely uncharacterized. To address this challenge, we developed a deep learning framework combining bulk and single-cell epigenomic data to evaluate the regulatory potential (i.e., silencing and activating strength) of noncoding AD variants in the dorsolateral prefrontal cortex (DLPFCs) and its major cell types. This model identified 1,457 silencer and 3,084 enhancer AD-associated variants in the DLPFC and binned them into silencer variants only (SL), enhancer variants only (EN), or both variant types (ENSL) classes. Each class exerts distinct cellular and molecular influences on AD pathogenesis. EN loci predominantly regulate housekeeping metabolic processes, whereas SL loci (including the genes MS4A6A , TREM2 , USP6NL , HLA-D ) are selectively linked to immune responses. Notably, 71% of these genes are significantly upregulated in AD and pro-inflammation-stimulated microglia. Furthermore, genes associated with SL loci are, in neuronal cells, often responsive to glutamate receptor antagonists (e.g, NBQX) and anti-inflammatory perturbagens (such as D-64131), the compound classes known for reducing the AD risk. ENSL loci, in contrast, are uniquely implicated in memory maintenance, neurofibrillary tangle assembly, and are also shared by other neurological disorders such as Parkinson's disease and schizophrenia. Key genes in this class of loci, such as MAPT , CR1/2 , and CLU , are frequently upregulated in AD subtypes with hyperphosphorylated tau aggregates. Critically, our model can accurately predict the impact of regulatory variants, with an average Pearson correlation coefficient of 0.54 and a directional concordance rate of 70% between our predictions and experimental outcomes. This model identified rs636317 as a causal AD variant in the MS4A locus, distinguishing it from the 7bp-away allele-neutral variant rs636341. Similarly, rs7922621 was prioritized over its 54-bp-away allele-neutral rs7901634 in the TSPAN14 locus. Additional causal variants include rs6701713 in the CR1 locus, and rs28834970 and rs755951 in the PTK2B locus. Collectively, this work advances our understanding of the regulatory landscape of AD-associated genetic variants, providing a framework to explore their functional roles in the pathogenesis of this complex disease.

PMID:40297423 | PMC:PMC12036408 | DOI:10.1101/2025.04.07.25325386

Digit Health. 2025 Apr 15;11:20552076251335542. doi: 10.1177/20552076251335542. eCollection 2025 Jan-Dec.

ABSTRACT

BACKGROUND: In modern workplaces, emotional well-being significantly impacts productivity, interpersonal relationships, and organizational stability. This study introduced an innovative facial-based emotion recognition system aimed at the real-time monitoring and management of employee emotional states.

METHODS: Utilizing the RetinaFace model for facial detection, the Dlib algorithm for feature extraction, and VGG16 for micro-expression classification, the system constructed a 10-dimensional emotion feature vector. Emotional anomalies were identified using the K-Nearest Neighbors algorithm and assessed with a 3σ-based risk evaluation method.

RESULTS: The system achieved high accuracy in emotion classification, as demonstrated by an empirical analysis, where VGG16 outperformed MobileNet and ResNet50 in key metrics such as accuracy, precision, and recall. Data augmentation techniques were employed to enhance the performance of the micro-expression classification model.

CONCLUSION: These techniques improved the across diverse emotional expressions, resulting in more accurate and robust emotion recognition. When deployed in a corporate environment, the system successfully monitored employees' emotional trends, identified potential risks, and provided actionable insights into early intervention. This study contributes to advancing corporate mental health management and lays the foundation for scalable emotion-based support systems in organizational settings.

PMID:40297378 | PMC:PMC12035250 | DOI:10.1177/20552076251335542

Digit Health. 2025 Apr 22;11:20552076251337183. doi: 10.1177/20552076251337183. eCollection 2025 Jan-Dec.

ABSTRACT

INTRODUCTION: The progression of Alzheimer's disease (AD) has been shown to significantly correlate with changes in brain tissue structure and leads to cognitive decline and dementia. Using radiomic features derived from brain magnetic resonance imaging (MRI) scan, we can get the help of deep learning (DL) model for diagnosing AD.

METHODS: This study proposes the use of the DL model under the framework of MR radiomics for AD diagnosis. Two cross-racial independent cohorts from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database (141 AD, 166 Mild Cognitive Impairment (MCI), and 231 normal control (NC) subjects) and Huashan hospital (45 AD, 35 MCI, and 31 NC subjects) were enrolled. We first performed preprocessing of MRI using methods such as spatial normalization and denoizing filtering. Next, we conducted Statistical Parametric Mapping analysis based on a two-sample t-test to identify regions of interest and extracted radiomic features using Radiomics tools. Subsequently, feature selection was carried out using the Least Absolute Shrinkage and Selection Operator model. Finally, the selected radiomic features were used to implement the AD diagnosis task with the TabNet model.

RESULTS: The model was quantitatively evaluated using the average values obtained from five-fold cross-validation. In the three-way classification task, the model achieved classification average area under the curve (AUC) of 0.8728 and average accuracy (ACC) of 0.7111 for AD versus MCI versus NC. For the binary classification task, the average AUC values were 0.8778, 0.8864, and 0.9506 for AD versus MCI, MCI versus NC, and AD versus NC, respectively, with average ACC of 0.8667, 0.8556, and 0.9222 for these comparisons.

CONCLUSIONS: The proposed model exhibited excellent performance in the AD diagnosis task, accurately distinguishing different stages of AD. This confirms the value of MR DL radiomic model for AD diagnosis.

PMID:40297370 | PMC:PMC12035500 | DOI:10.1177/20552076251337183

Digit Health. 2025 Apr 15;11:20552076251333660. doi: 10.1177/20552076251333660. eCollection 2025 Jan-Dec.

ABSTRACT

OBJECTIVE: This study aims to develop and validate a deep learning model to predict smoking status in patients with chronic obstructive pulmonary disease (COPD) using data from a national survey.

METHODS: Data from the Korea National Health and Nutrition Examination Survey (2007-2018) were used to extract 5466 COPD-eligible cases. The data collection involved demographic, behavioral, and clinical variables, including 21 predictors such as age, sex, and pulmonary function test results. The dependent variable, smoking status, was categorized as smoker or nonsmoker. A residual neural network (ResNN) model was developed and compared with five machine learning algorithms (random forest, decision tree, Gaussian Naive Bayes, K-nearest neighbor, and AdaBoost) and two deep learning models (multilayer perceptron and TabNet). Internal validation was performed using five-fold cross-validation, and model performance was evaluated using the area under the receiver operating characteristic (AUROC) curve, sensitivity, specificity, and F1-score.

RESULTS: The ResNN achieved an AUROC, sensitivity, specificity, and F1-score of 0.73, 70.1%, 75.2%, and 0.67, respectively, outperforming previous machine learning and deep learning models in predicting smoking status in patients with COPD. Explainable artificial intelligence (Shapley additive explanations) identified key predictors, including sex, age, and perceived health status.

CONCLUSION: This deep learning model accurately predicts smoking status in patients with COPD, offering potential as a decision-support tool to detect high-risk persistent smokers for targeted interventions. Future studies should focus on external validation and incorporate additional behavioral and psychological variables to improve its generalizability and performance.

PMID:40297369 | PMC:PMC12035114 | DOI:10.1177/20552076251333660

Digit Health. 2025 Apr 15;11:20552076251334431. doi: 10.1177/20552076251334431. eCollection 2025 Jan-Dec.

ABSTRACT

BACKGROUND: When applying for teleconsultations, medical laboratory reports are usually photographed with a mobile phone, and the photographic results are uploaded as teleconsultation application materials. It is very meaningful to extract the content of the image medical laboratory report and store the content digitally. There are already applications of OCR technology for medical text file recognition, but no researchers have recognized the format of the medical laboratory report and obtained the report content as a serialized process to digitize the image report. This article proposes a serialization method to digitize the medical laboratory report image.

MATERIALS AND METHODS: This article first collects 330 image-based medical laboratory reports, annotates the format of the medical laboratory reports, and forms a training dataset for the layout analysis model. Then, using the pre-trained model, the dataset is trained to obtain a layout analysis model that can correctly recognize the format of the medical laboratory report. Then, the layout of the input image-based medical laboratory report is analyzed, and the layout analysis results are used to call the text detection and text recognition models to obtain the digital content of the image report. Finally, adjusting the layout of the digital content and storing the digital content as a docx file.

RESULTS: After training the layout analysis model, integrating layout analysis, text detection, and text recognition, we have obtained a serialization method that digitizes the content of the image medical laboratory report, restores the report format, shields sensitive and irrelevant content, and digitizes the report content of interest.

CONCLUSIONS: By digitizing the image medical laboratory report through the serialization method, we can correctly display the content of the medical laboratory report for teleconsultation, while removing irrelevant content in the report, such as user names, examination equipment numbers, etc.

PMID:40297365 | PMC:PMC12035204 | DOI:10.1177/20552076251334431

Digit Health. 2025 Apr 15;11:20552076251334040. doi: 10.1177/20552076251334040. eCollection 2025 Jan-Dec.

ABSTRACT

BACKGROUND: Rapid and accurate identification of large vessel occlusion (LVO) is crucial for determining eligibility for endovascular treatment. We aimed to validate whether computed tomography combined with clinical information (CT&CI) or diffusion-weighted imaging (DWI) offers better predictive accuracy for anterior circulation LVO.

METHODS: Computed tomography combined with clinical information and DWI data from patients diagnosed with acute ischemic stroke were collected. Three deep-learning models, convolutional neural network, EfficientNet-B2, and DenseNet121, were used to compare CT&CI and DWI for detecting anterior circulation LVO.

RESULTS: A total of 456 patients, 228 patients with LVO [68.91 ± 12.84 years, 63.60% male; initial National Institutes of Health Stroke Scale (NIHSS) score: median 11 (7-14)] and without LVO [67.06 ± 12.29 years, 64.04% male; initial NIHSS score: median 2 (1-4)] were enrolled. Diffusion-weighted imaging achieved better results than CT&CI did in each performance metric. In DenseNet121, the area under the curves (AUCs) were found to be 0.833 and 0.756, respectively, while in EfficientNet-B2, the AUCs were 0.815 and 0.647, respectively.

CONCLUSIONS: In detecting the presence of anterior circulation LVO, DWI showed better results in each performance metric than CT&CI did, and the best-performing deep-learning model was DenseNet121.

PMID:40297357 | PMC:PMC12035260 | DOI:10.1177/20552076251334040

Front Mol Biosci. 2025 Apr 14;12:1561464. doi: 10.3389/fmolb.2025.1561464. eCollection 2025.

ABSTRACT

Breast cancer (BC) remains a complex and widespread problem, affecting millions of women worldwide, Among the various subtypes of BC, triple-negative breast cancer (TNBC) is particularly challenging, representing approximately 20% of all BC cases, and the survival rate of TNBC patients is generally worse than other subtypes of BC. TNBC is a heterogeneous disease characterized by lack of expression of three receptors: estrogen (ER), progesterone (PR), and human epidermal growth factor receptor 2 (HER2), resulting conventional hormonal therapies are ineffective for its management. Despite various therapeutic approaches have been explored, but no definitive solution has been found yet for TNBC. Current treatments options are chemotherapy, immunotherapy, radiotherapy and surgery, although, these therapies have some limitations, such as the development of resistance to anti-cancer drugs, and off-target toxicity, which remain primary obstacles and significant challenges for TNBC. Several findings have shown that EVs exhibit significant therapeutic promise in many diseases, and a similar important role has been observed in various types of tumor. Studies suggest that EVs may offer a potential solution for the management of TNBC. This review highlights the multifaceted roles of EVs in TNBC, emphasizing their involvement in disease progression, diagnosis and therapeutic approach, as well as their potential as biomarkers and drug delivery.

PMID:40297849 | PMC:PMC12034555 | DOI:10.3389/fmolb.2025.1561464

Front Physiol. 2025 Apr 14;16:1535331. doi: 10.3389/fphys.2025.1535331. eCollection 2025.

ABSTRACT

BACKGROUND: Assessing real-time stress in individuals to prevent the accumulation of stress is necessary due to the adverse effects of excessive psychological stress on health. Since both stress and circadian rhythms affect the excitability of the nervous system, the influence of circadian rhythms needs to be considered during stress assessment. Most studies train classifiers using physiological data collected during fixed short time periods, overlooking the assessment of stress levels at other times.

METHODS: In this work, we propose a method for training a classifier capable of identifying stress and resting states throughout the day, based on 10 short-term heart rate variability (HRV) feature data obtained from morning, noon, and evening. To characterize the circadian rhythms of HRV features, heartbeat interval data were collected and analyzed from 50 volunteers over three consecutive days. The circadian rhythm trends in the HRV features were then removed using the Smoothness Priors Approach (SPA), and XGBoost models were trained to assess stress.

RESULTS: The results show that all HRV features exhibit 12-h and 24-h circadian rhythms, and the circadian rhythm differences across different days for individuals are relatively small. Furthermore, training classifiers on detrended data can improve the overall accuracy of stress assessment across all time periods. Specifically, when combining data from different time periods as the training dataset, the accuracy of the classifier trained on detrended data increases by 13.67%.

DISCUSSION: These findings indicate that using HRV features with circadian rhythm trends removed is an effective method for assessing stress at all times throughout the day.

PMID:40297780 | PMC:PMC12034550 | DOI:10.3389/fphys.2025.1535331