BMC Pulm Med. 2025 Jan 29;25(1):47. doi: 10.1186/s12890-025-03520-4.

ABSTRACT

BACKGROUND: Fibrotic types of interstitial lung abnormalities seen on high-resolution computed tomography scans, characterised by traction bronchiolectasis/bronchiectasis with or without honeycombing, are predictors of progression and poor prognostic factors of interstitial lung abnormalities. There are no reports on the clinical characteristics of fibrotic interstitial lung abnormalities on high-resolution computed tomography scans. Therefore, we aimed to examine these clinical characteristics and clarify the predictive factors of fibrotic interstitial lung abnormalities on high-resolution computed tomography scans.

METHODS: Clinical and paraclinical data of 164 patients enrolled in the initial year of a multicentre prospective observational study (Kumamoto interstitial lung abnormalities study in Japan) involving over 62,000 examinees during routine health examinations were analysed. Clinical laboratory evaluations are expressed as medians and interquartile ranges for each evaluation time point, and boxplots were created for graphical representation. The percentages of abnormal clinical laboratory results were compared between the groups using chi-square or Fisher's exact tests. Univariate or multivariate logistic regression analyses were performed to analyse the relationship between fibrotic interstitial lung abnormalities and other clinical factors.

RESULTS: Fibrotic interstitial lung abnormalities were observed on high-resolution computed tomography scans in 135 (82%) patients at the time of diagnosis. Multivariate analysis showed that older age (Odds ratio, 1.06; 95% confidence interval, 1.01-1.12; p = 0.021), auscultatory fine crackles (Odds ratio, 3.39; 95% confidence interval, 1.33-8.65; p < 0.01), and elevated serum surfactant protein-D (Odds ratio, 2.68; 95% confidence interval, 1.02-8.64; p = 0.045) were independent predictive factors of fibrotic interstitial lung abnormalities. The predicted area under the curve of the fibrotic interstitial lung abnormalities based on these three factors was 0.77 (95% confidence interval, 0.68-0.86). The proportion of undecided diagnoses in the fibrotic interstitial lung abnormalities group (14%) was significantly lower than that in the non-fibrotic interstitial lung abnormalities group (41%) (p = 0.0027).

CONCLUSIONS: Fine crackles on auscultation and elevated serum surfactant protein-D levels are predictors of fibrotic interstitial lung abnormalities in older patients with interstitial lung abnormalities. These findings may assist non-radiological physicians in referring patients to specialists for early intervention in progressive fibrotic interstitial lung diseases.

TRIAL REGISTRATION NUMBER/DATE: UMIN000045149/2021.12.1.

PMID:39881354 | DOI:10.1186/s12890-025-03520-4

Biochem Genet. 2025 Jan 29. doi: 10.1007/s10528-024-11012-z. Online ahead of print.

ABSTRACT

Idiopathic pulmonary fibrosis (IPF) is a rapidly progressive interstitial lung disease of unknown pathogenesis with no effective treatment currently available. Given the regulatory roles of lncRNAs (TP53TG1, LINC00342, H19, MALAT1, DNM3OS, MEG3), miRNAs (miR-218-5p, miR-126-3p, miR-200a-3p, miR-18a-5p, miR-29a-3p), and their target protein-coding genes (PTEN, TGFB2, FOXO3, KEAP1) in the TGF-β/SMAD3, Wnt/β-catenin, focal adhesion, and PI3K/AKT signaling pathways, we investigated the expression levels of selected genes in peripheral blood mononuclear cells (PBMCs) and lung tissue from patients with IPF. Lung tissue and blood samples were collected from 33 newly diagnosed, treatment-naive patients and 70 healthy controls. Gene expression levels were analyzed by RT-qPCR. TaqMan assays and TaqMan MicroRNA assay were employed to quantify the expression of target lncRNAs, mRNAs, and miRNAs. Our study identified significant differential expression in PBMCs from IPF patients compared to healthy controls, including lncRNAs MALAT1 (Fold Change = 3.809, P = 0.0001), TP53TG1 (Fold Change = 0.4261, P = 0.0021), and LINC00342 (Fold Change = 1.837, P = 0.0448); miRNAs miR-126-3p (Fold Change = 0.102, P = 0.0028), miR-200a-3p (Fold Change = 0.442, P = 0.0055), and miR-18a-5p (Fold Change = 0.154, P = 0.0034); and mRNAs FOXO3 (Fold Change = 4.604, P = 0.0032) and PTEN (Fold Change = 2.22, P = 0.0011). In lung tissue from IPF patients, significant expression changes were observed in TP53TG1 (Fold Change = 0.2091, P = 0.0305) and DNM3OS (Fold Change = 4.759, P = 0.05). Combined analysis of PBMCs expression levels for TP53TG1, MALAT1, miRNA miR-126-3p, and PTEN distinguished IPF patients from healthy controls with an AUC = 0.971, sensitivity = 0.80, and specificity = 0.955 (P = 6 × 10-8). These findings suggest a potential involvement of the identified ncRNAs and mRNAs in IPF pathogenesis. However, additional functional validation studies are needed to elucidate the precise molecular mechanisms by which these lncRNAs, miRNAs, and their targets contribute to PF.

PMID:39881079 | DOI:10.1007/s10528-024-11012-z

Environ Toxicol Chem. 2025 Jan 28:vgaf028. doi: 10.1093/etojnl/vgaf028. Online ahead of print.

ABSTRACT

Given the need to reduce animal testing for environmental risk assessment, we aim to develop a fish invitrome, an alternative fish modular framework capable of predicting chemical toxicity in fish without the use of animals. The central module of the framework is the validated RTgill-W1 cell line assay that predicts fish acute toxicity of chemicals (Organization for Economic Cooperation and Development Test Guideline (OECD TG) 249). Expanding towards prediction of chronic toxicity, the fish invitrome includes two other well-advanced modules for chemical bioaccumulation/biotransformation and inhibition of fish growth. This framework is expected to continuously evolve with the development of modules that predict, for instance, neurotoxicity and reproductive toxicity. We envisage the fish invitrome framework to become part of the broader academic field of New Approach Methodologies (NAMs), where it will remain flexible and open to integration of new developments from research groups around the world. To accelerate the development and uptake of this framework, we strive for transdisciplinarity, integrating both natural and social sciences, along with broader stakeholder interactions. A stepwise socio-technical approach has been chosen, where mainstreaming the fish invitrome involves progressive adoption across various ecotoxicological contexts. The framework will be co-designed with stakeholders from academia, industry, and regulatory bodies. Rather than aiming for immediate regulatory acceptance, this approach aims to build trust and familiarity with fish cell line-based testing among stakeholders. By doing so, it encourages broader use of the framework in practical applications while gradually overcoming institutional, cultural, and technical barriers. Additionally, establishing a clear roadmap for mainstreaming the fish invitrome will help identify and address challenges to its uptake, ensuring a smoother transition to non-organismal testing methodologies.

PMID:39880375 | DOI:10.1093/etojnl/vgaf028

Biochim Biophys Acta Gen Subj. 2025 Jan 27:130768. doi: 10.1016/j.bbagen.2025.130768. Online ahead of print.

ABSTRACT

Apolipoprotein E (apoE) polymorphism is associated with different pathologies such as atherosclerosis and Alzheimer's disease. Knowledge of the three-dimensional structure of apoE and isoform-specific structural differences are prerequisites for the rational design of small molecule structure modulators that correct the detrimental effects of pathological isoforms. In this study, cross-linking mass spectrometry (XL-MS) targeting Asp, Glu and Lys residues was used to explore the intramolecular interactions in the E2, E3 and E4 isoforms of apoE. The resulting quantitative XL-MS data combined with molecular modeling revealed isoform-specific characteristics of the N- and C-terminal domain interfaces as well as the isoform-dependent dynamic equilibrium of these interfaces. Finally, the data identified a network of salt bridges formed by R61-R112-E109 residues in the N-terminal helical bundle as a modulator of the interaction with the C-terminal domain making this network a potential drug target.

PMID:39880049 | DOI:10.1016/j.bbagen.2025.130768

Curr Biol. 2025 Jan 23:S0960-9822(24)01707-X. doi: 10.1016/j.cub.2024.12.039. Online ahead of print.

ABSTRACT

Yeasts are a diverse group of unicellular fungi that have developed a wide array of phenotypes and traits over 400 million years of evolution. However, we still lack an understanding of the biological principles governing the range of cell morphologies, metabolic modes, and reproductive strategies yeasts display. In this study, we explored the relationship between cell morphology and metabolism in sixteen yeast strains across eleven species. We performed a quantitative analysis of the physiology and morphology of these strains and discovered a strong correlation between the glucose uptake rate (GUR) and the surface-area-to-volume ratio. 14C-glucose uptake experiments demonstrated that the GUR for a given strain is governed either by glucose transport capacity or glycolytic rate, indicating that it is rather the rate of glucose metabolism in general that correlates with cell morphology. Furthermore, perturbations in glucose metabolism influenced cell sizes, whereas manipulating cell size did not affect GUR, suggesting that glucose metabolism determines cell size rather than the reverse. Across the strains tested, we also found that the rate of glucose metabolism influenced ethanol production rate, biomass yield, and carbon dioxide transfer rate. Overall, our findings demonstrate that the rate of glucose metabolism is a key factor shaping yeast cell morphology and physiology, offering new insights into the fundamental principles of yeast biology.

PMID:39879976 | DOI:10.1016/j.cub.2024.12.039

Mar Pollut Bull. 2025 Jan 28;212:117615. doi: 10.1016/j.marpolbul.2025.117615. Online ahead of print.

ABSTRACT

The Strait of Sicily, a vital marine passage with diverse fauna, is seeing a steep rise in the planning of offshore wind farm projects. This study assesses the acoustic impact of these wind farms on local marine species. Underwater propagation was modeled for three proposed floating wind farms using JASCO's Marine Operations Noise Model (MONM), which integrates a parabolic equation method for frequencies from 10 to 800 Hz and a beam-tracing model for 1 to 25 kHz. Propagation losses were calculated in one-third octave bands for ten source locations selected to represent the variability in bathymetry, and considering sound speed profiles for February and August. Sound levels from floating turbines were used to estimate exceedance ranges to known acoustic thresholds for marine species. Modeling indicated that sound levels could exceed temporary threshold shift and, for some species, permanent threshold shift criteria within a few tens of meters, but only if animals were to remain for 24 h at such small distances from a turbine. Behavioral disturbance thresholds for marine mammals were exceeded up to 68 km from the wind farms' boundaries. The study emphasizes considering species-specific sensitivities and ecological contexts in environmental impact assessments, recommending mitigation measures, such as the strategic placement of the turbines and continuous monitoring, to minimize adverse effects on local marine fauna, including marine mammals and turtles.

PMID:39879849 | DOI:10.1016/j.marpolbul.2025.117615

Stem Cell Res. 2025 Jan 15;83:103660. doi: 10.1016/j.scr.2025.103660. Online ahead of print.

ABSTRACT

NHD/PLOSL is an orphan disease characterized by progressive presenile dementia associated with recurrent fractures due to polycystic bone lesions. In this study, we generated the human induced pluripotent stem cell (hiPSC) line BIHi292-A from a 30-year-old women diagnosed with NHD/PLOSL, carrying two compound heterozygous frameshift mutations [c.313del (p.Ala105fs) and c.199del (p.His67fs)] in the TREM2 (triggering receptor expressed on myeloid cells 2) gene. BIHi292-A hiPSCs are karyotypically normal, express typical markers for the undifferentiated state and have pluripotent differentiation potential. BIHi292-A cells will provide a valuable tool for investigating pathogenic mechanisms of NHD/PLOSL and TREM2-related research questions.

PMID:39879812 | DOI:10.1016/j.scr.2025.103660

Farm Hosp. 2025 Jan 28:S1130-6343(24)00154-5. doi: 10.1016/j.farma.2024.09.005. Online ahead of print.

ABSTRACT

BACKGROUND: Adalimumab biosimilar MSB11022 (Idacio®) has been approved for the same indications as its originator (Humira®), based on findings from clinical trials in plaque psoriasis. Data on its efficacy and safety in inflammatory bowel disease, however, are scarce.

METHODS: Retrospective, observational study of 44 patients with inflammatory bowel disease: 30 were treated with originator adalimumab, five were directly started on MSB11022, and nine switched from originator to biosimilar adalimumab. To evaluate the effectiveness of the use of adalimumab in inflammatory bowel disease, both laboratory markers (fecal calprotectin and C-reactive protein) and scales that measure the activity of inflammatory bowel disease using specific scales (Harvey-Bradshaw Index [HBI] have been used for Crohn's disease and Mayo Score for ulcerative colitis). Efficacy was evaluated by recording the adverse effects that could occur with the administration of adalimumab (original or biosimilar). The success of the switch was determined by analyzing meaningful differences in effectiveness and safety criteria. Concomitant therapy and the need for dose intensification were also analyzed. Objective of this study was to assess the effectiveness and safety of biosimilar adalimumab in adalimumab-naïve patients and patients switched from originator adalimumab.

RESULTS: No significant differences were observed in clinical disease activity (p = 0.317) or biochemical parameters (fecal calprotectin [p = 0.445] and C-reactive protein [p = 0.661]) after the switch from the originator adalimumab to MSB11022. There was not a significant reduction in the concomitant use of corticosteroids and thiopurines (p = 0.157). No emergency room visits or hospitalizations were observed during the study period and none of the patients experienced serious adverse effects.

CONCLUSIONS: Between originator adalimumab and biosimilar-start cohorts no differences were observed, between originator adalimumab and switch cohorts no significant differences were found either, and with the pre- and post-switch to biosimilar comparison two of the nine patients experienced adverse effects after the switch. The biosimilar showed a favorable safety profile (one patient with a serious adverse effect [rash] with biosimilar discontinued treatment) and no significant changes to clinical or biochemical parameters were observed after the switch.

PMID:39880784 | DOI:10.1016/j.farma.2024.09.005

Arab J Gastroenterol. 2025 Jan 28:S1687-1979(25)00004-8. doi: 10.1016/j.ajg.2025.01.004. Online ahead of print.

ABSTRACT

BACKGROUND AND STUDY AIMS: Non-alcoholic fatty liver disease (NAFLD) is the most prevalent chronic liver disorder, affecting 23% to 32% of the global population. This clinical study aimed to assess the efficacy of Curcuma zedoaria Rosc. compared to vitamin E in managing NAFLD.

PATIENTS AND METHODS: In this randomized, single-blind, standard-controlled study, 68 patients with grade 1 (mild) and grade 2 (moderate) NAFLD were randomly assigned to receive either Curcuma zedoaria Rosc. powder in capsule form (500 mg orally, twice a day) in the test group or vitamin E (400 mg orally, twice a day) in the control group for 60 days. Secondary endpoints included improvements in fatty liver grades, ultra-sonographic liver span, lipid profile, and liver function parameters after 60 days. Primary endpoints included improvements in dull ache intensity in the right hypochondrium (RHC), dyspepsia, anorexia, and severity of malaise assessed at days 0, 15, 30, 45, and 61.

RESULTS: Per protocol analysis was performed on 50 patients who completed the study. Both test and control groups showed significant improvement in dull ache severity in the RHC (p < 0.0001). The test group exhibited more favorable outcomes post-treatment (Chi-sq = 23.17, df = 2, p < 0.0001). Dyspepsia severity significantly improved in both groups post-treatment (p = 0.005 and p = 0.010, respectively), with the test group showing slightly better outcomes. Anorexia significantly improved in the test group (p = 0.016) from 72.00 % reporting absence post-treatment to 100.00 % absence, while the control group showed improvement without statistical significance (p = 0.102). Malaise severity significantly improved in the test group (p < 0.0001), with 84.00 % reporting absence post-treatment compared to 8.00 % in the control group, showing significant differences (p < 0.0001). Both groups exhibited a significant reduction in liver span post-treatment (p-value < 0.0001) without inter- group differences. Fatty liver grades improved significantly in both groups post-treatment (p < 0.0001), with no significant difference between groups (Chi-sq = 4, df = 2, p = 0.1353). There were no changes in liver function markers and lipid parameters in both groups, though the test drug demonstrated a slight reduction in serum triglyceride levels. No drug-related adverse events were observed during the trial.

CONCLUSION: The study revealed that Curcuma zedoaria Rosc. is effective in managing NAFLD, showing better outcomes than vitamin E in subjective parameters like dyspepsia, malaise, anorexia, and dull ache in RHC. With no observed drug-related adverse events, Curcuma zedoaria Rosc. could be a suitable alternative to conventional treatment modalities for NAFLD.

PMID:39880723 | DOI:10.1016/j.ajg.2025.01.004

JMIR Res Protoc. 2025 Jan 29;14:e63597. doi: 10.2196/63597.

ABSTRACT

BACKGROUND: TheKeep.Ca was built to facilitate engagement with those experiencing cancer in Manitoba, Canada. Constructed between 2020 and 2024 with a group of patient advisors, the website includes information on engagement activities including research participation, the patient advisor role, and how those experiencing cancer can access these Manitoba activities. A link allows visitors to register to be contacted about activities that match their demographics, cancer history, and activity preferences. After TheKeep.Ca was constructed, this protocol was developed to establish TheKeep.Ca as a platform for scientific research focused on optimally engaging those experiencing cancer.

OBJECTIVE: We asked the following questions: (1) What was the patient advisors' experience who participated in developing TheKeep.Ca? (2) What are the baseline characteristics of website traffic and registrants at TheKeep.Ca? (3) How does registering with TheKeep.Ca impact the cancer experience?

METHODS: The planned launch date for the website and initiation of research activities is January 2025. For objective 1, the active patient advisors (N=6) participating in the website project will be invited to participate in project activities including with responses to a question prompt sheet, semistructured audio-recorded interviews, or both. Responses and interviews will be analyzed using reflexive thematic analysis to understand and inform practices for patient engagement on projects. At the website launch, TheKeep.Ca will become publicly accessible and indexable on internet search engines, but no additional promotional interventions will take place in the initial 6 months resulting in visitors primarily from web search traffic. For objective 2, Google Analytics and website registrant data collected during the first six months will be analyzed to obtain baseline characteristics of website visitors. For objective 3, an online survey will be emailed to registrants six months after the website launch characterizing their website experience, the activities they participated in, and collecting feedback on the website. For objectives 2 and 3, quantitative data will be analyzed using both descriptive and inferential statistics, and qualitative data from open-ended questions will be analyzed using thematic analysis guided by an inductive descriptive semantic approach.

RESULTS: This study was approved by the University of Manitoba Health Research Ethics Board on December 12, 2024 (HS26614-H2024L263). Institutional approval from CancerCare Manitoba is pending as of December 23, 2024. Findings from objective 1 are expected to be finalized within the first six months after the website launch. Those from objectives 2 and 3 are expected by the 12-month mark. Reporting will include peer-reviewed journals, conferences, and a lay-language summary on TheKeep.Ca.

CONCLUSIONS: The research outlined in this protocol will facilitate understanding patient advisors' experience in developing TheKeep.Ca. It will also characterize the website' effectiveness and its impact on the cancer experience, providing a baseline and direction for future research and development.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/63597.

PMID:39879620 | DOI:10.2196/63597

J Bras Pneumol. 2025 Jan 27;51(1):e20240258. doi: 10.36416/1806-3756/e20240258.

NO ABSTRACT

PMID:39879516 | DOI:10.36416/1806-3756/e20240258

Appl Neuropsychol Adult. 2025 Jan 29:1-10. doi: 10.1080/23279095.2025.2457999. Online ahead of print.

ABSTRACT

In the fields of engineering, science, technology, and medicine, artificial intelligence (AI) has made significant advancements. In particular, the application of AI techniques in medicine, such as machine learning (ML) and deep learning (DL), is rapidly growing and offers great potential for aiding physicians in the early diagnosis of illnesses. Depression, one of the most prevalent and debilitating mental illnesses, is projected to become the leading cause of disability worldwide by 2040. For early diagnosis, a patient-friendly, cost-effective approach based on readily observable and objective indicators is essential. The objective of this research is to develop machine learning and deep learning techniques that utilize electroencephalogram (EEG) signals to diagnose depression. Different statistical features were extracted from the EEG signals and fed into the models. Three classifiers were constructed: 1D Convolutional Neural Network (1DCNN), Support Vector Machine (SVM), and Logistic Regression (LR). The methods were tested on a dataset comprising EEG signals from 34 patients with Major Depressive Disorder (MDD) and 30 healthy subjects. The signals were collected under three distinct conditions: TASK, when the subject was performing a task; Eye Close (EC), when the subject's eyes were closed; and Eye Open (EO), when the subject's eyes were open. All three classifiers were applied to each of the three types of signals, resulting in nine (3 × 3) experiments. The results showed that TASK signals yielded the highest accuracies of 88.4%, 89.3%, and 90.21% for LR, SVM, and 1DCNN, respectively, compared to EC and EO signals. Additionally, the proposed methods outperformed some state-of-the-art approaches. These findings highlight the potential of EEG-based approaches for the clinical diagnosis of depression and provide promising avenues for further research. Additionally, the proposed methodology demonstrated statistically significant improvements in classification accuracy, with p-values < 0.05, ensuring robustness and reliability.

PMID:39879638 | DOI:10.1080/23279095.2025.2457999

J Med Internet Res. 2025 Jan 29;27:e57723. doi: 10.2196/57723.

ABSTRACT

BACKGROUND: Neuroimaging segmentation is increasingly important for diagnosing and planning treatments for neurological diseases. Manual segmentation is time-consuming, apart from being prone to human error and variability. Transformers are a promising deep learning approach for automated medical image segmentation.

OBJECTIVE: This scoping review will synthesize current literature and assess the use of various transformer models for neuroimaging segmentation.

METHODS: A systematic search in major databases, including Scopus, IEEE Xplore, PubMed, and ACM Digital Library, was carried out for studies applying transformers to neuroimaging segmentation problems from 2019 through 2023. The inclusion criteria allow only for peer-reviewed journal papers and conference papers focused on transformer-based segmentation of human brain imaging data. Excluded are the studies dealing with nonneuroimaging data or raw brain signals and electroencephalogram data. Data extraction was performed to identify key study details, including image modalities, datasets, neurological conditions, transformer models, and evaluation metrics. Results were synthesized using a narrative approach.

RESULTS: Of the 1246 publications identified, 67 (5.38%) met the inclusion criteria. Half of all included studies were published in 2022, and more than two-thirds used transformers for segmenting brain tumors. The most common imaging modality was magnetic resonance imaging (n=59, 88.06%), while the most frequently used dataset was brain tumor segmentation dataset (n=39, 58.21%). 3D transformer models (n=42, 62.69%) were more prevalent than their 2D counterparts. The most developed were those of hybrid convolutional neural network-transformer architectures (n=57, 85.07%), where the vision transformer is the most frequently used type of transformer (n=37, 55.22%). The most frequent evaluation metric was the Dice score (n=63, 94.03%). Studies generally reported increased segmentation accuracy and the ability to model both local and global features in brain images.

CONCLUSIONS: This review represents the recent increase in the adoption of transformers for neuroimaging segmentation, particularly for brain tumor detection. Currently, hybrid convolutional neural network-transformer architectures achieve state-of-the-art performances on benchmark datasets over standalone models. Nevertheless, their applicability remains highly limited by high computational costs and potential overfitting on small datasets. The heavy reliance of the field on the brain tumor segmentation dataset hints at the use of a more diverse set of datasets to validate the performances of models on a variety of neurological diseases. Further research is needed to define the optimal transformer architectures and training methods for clinical applications. Continuing development may make transformers the state-of-the-art for fast, accurate, and reliable brain magnetic resonance imaging segmentation, which could lead to improved clinical tools for diagnosing and evaluating neurological disorders.

PMID:39879621 | DOI:10.2196/57723

JMIR Public Health Surveill. 2025 Jan 29;11:e63809. doi: 10.2196/63809.

ABSTRACT

BACKGROUND: Suicide represents a critical public health concern, and machine learning (ML) models offer the potential for identifying at-risk individuals. Recent studies using benchmark datasets and real-world social media data have demonstrated the capability of pretrained large language models in predicting suicidal ideation and behaviors (SIB) in speech and text.

OBJECTIVE: This study aimed to (1) develop and implement ML methods for predicting SIBs in a real-world crisis helpline dataset, using transformer-based pretrained models as a foundation; (2) evaluate, cross-validate, and benchmark the model against traditional text classification approaches; and (3) train an explainable model to highlight relevant risk-associated features.

METHODS: We analyzed chat protocols from adolescents and young adults (aged 14-25 years) seeking assistance from a German crisis helpline. An ML model was developed using a transformer-based language model architecture with pretrained weights and long short-term memory layers. The model predicted suicidal ideation (SI) and advanced suicidal engagement (ASE), as indicated by composite Columbia-Suicide Severity Rating Scale scores. We compared model performance against a classical word-vector-based ML model. We subsequently computed discrimination, calibration, clinical utility, and explainability information using a Shapley Additive Explanations value-based post hoc estimation model.

RESULTS: The dataset comprised 1348 help-seeking encounters (1011 for training and 337 for testing). The transformer-based classifier achieved a macroaveraged area under the curve (AUC) receiver operating characteristic (ROC) of 0.89 (95% CI 0.81-0.91) and an overall accuracy of 0.79 (95% CI 0.73-0.99). This performance surpassed the word-vector-based baseline model (AUC-ROC=0.77, 95% CI 0.64-0.90; accuracy=0.61, 95% CI 0.61-0.80). The transformer model demonstrated excellent prediction for nonsuicidal sessions (AUC-ROC=0.96, 95% CI 0.96-0.99) and good prediction for SI and ASE, with AUC-ROCs of 0.85 (95% CI 0.97-0.86) and 0.87 (95% CI 0.81-0.88), respectively. The Brier Skill Score indicated a 44% improvement in classification performance over the baseline model. The Shapley Additive Explanations model identified language features predictive of SIBs, including self-reference, negation, expressions of low self-esteem, and absolutist language.

CONCLUSIONS: Neural networks using large language model-based transfer learning can accurately identify SI and ASE. The post hoc explainer model revealed language features associated with SI and ASE. Such models may potentially support clinical decision-making in suicide prevention services. Future research should explore multimodal input features and temporal aspects of suicide risk.

PMID:39879608 | DOI:10.2196/63809

JMIR Form Res. 2025 Jan 29;9:e66330. doi: 10.2196/66330.

ABSTRACT

BACKGROUND: Estimating the prevalence of schizophrenia in the general population remains a challenge worldwide, as well as in Japan. Few studies have estimated schizophrenia prevalence in the Japanese population and have often relied on reports from hospitals and self-reported physician diagnoses or typical schizophrenia symptoms. These approaches are likely to underestimate the true prevalence owing to stigma, poor insight, or lack of access to health care among respondents. To address these issues, we previously developed an artificial neural network (ANN)-based schizophrenia classification model (SZ classifier) using data from a large-scale Japanese web-based survey to enhance the comprehensiveness of schizophrenia case identification in the general population. In addition, we also plan to introduce a population-based survey to collect general information and sample participants matching the population's demographic structure, thereby achieving a precise estimate of the prevalence of schizophrenia in Japan.

OBJECTIVE: This study aimed to estimate the prevalence of schizophrenia by applying the SZ classifier to random samples from the Japanese population.

METHODS: We randomly selected a sample of 750 participants where the age, sex, and regional distributions were similar to Japan's demographic structure from a large-scale Japanese web-based survey. Demographic data, health-related backgrounds, physical comorbidities, psychiatric comorbidities, and social comorbidities were collected and applied to the SZ classifier, as this information was also used for developing the SZ classifier. The crude prevalence of schizophrenia was calculated through the proportion of positive cases detected by the SZ classifier. The crude estimate was further refined by excluding false-positive cases and including false-negative cases to determine the actual prevalence of schizophrenia.

RESULTS: Out of 750 participants, 62 were classified as schizophrenia cases by the SZ classifier, resulting in a crude prevalence of schizophrenia in the general population of Japan of 8.3% (95% CI 6.6%-10.1%). Among these 62 cases, 53 were presumed to be false positives, and 3 were presumed to be false negatives. After adjustment, the actual prevalence of schizophrenia in the general population was estimated to be 1.6% (95% CI 0.7%-2.5%).

CONCLUSIONS: This estimated prevalence was slightly higher than that reported in previous studies, possibly due to a more comprehensive disease classification methodology or, conversely, model limitations. This study demonstrates the capability of an ANN-based model to improve the estimation of schizophrenia prevalence in the general population, offering a novel approach to public health analysis.

PMID:39879582 | DOI:10.2196/66330

Brief Bioinform. 2024 Nov 22;26(1):bbaf039. doi: 10.1093/bib/bbaf039.

ABSTRACT

Pathway analysis plays a critical role in bioinformatics, enabling researchers to identify biological pathways associated with various conditions by analyzing gene expression data. However, the rise of large, multi-center datasets has highlighted limitations in traditional methods like Over-Representation Analysis (ORA) and Functional Class Scoring (FCS), which struggle with low signal-to-noise ratios (SNR) and large sample sizes. To tackle these challenges, we use a deep learning-based classification method, Gene PointNet, and a novel $P$-value computation approach leveraging the confusion matrix to address pathway analysis tasks. We validated our method effectiveness through a comparative study using a simulated dataset and RNA-Seq data from The Cancer Genome Atlas breast cancer dataset. Our method was benchmarked against traditional techniques (ORA, FCS), shallow machine learning models (logistic regression, support vector machine), and deep learning approaches (DeepHisCom, PASNet). The results demonstrate that GPNet outperforms these methods in low-SNR, large-sample datasets, where it remains robust and reliable, significantly reducing both Type I error and improving power. This makes our method well suited for pathway analysis in large, multi-center studies. The code can be found at https://github.com/haolu123/GPNet_pathway">https://github.com/haolu123/GPNet_pathway.

PMID:39879387 | DOI:10.1093/bib/bbaf039

Brief Bioinform. 2024 Nov 22;26(1):bbaf037. doi: 10.1093/bib/bbaf037.

ABSTRACT

This study aimed to investigate the genetic association between glioblastoma (GBM) and unsupervised deep learning-derived imaging phenotypes (UDIPs). We employed a combination of genome-wide association study (GWAS) data, single-nucleus RNA sequencing (snRNA-seq), and scPagwas (pathway-based polygenic regression framework) methods to explore the genetic links between UDIPs and GBM. Two-sample Mendelian randomization analyses were conducted to identify causal relationships between UDIPs and GBM. Colocalization analysis was performed to validate genetic associations, while scPagwas analysis was used to evaluate the relevance of key UDIPs to GBM at the cellular level. Among 512 UDIPs tested, 23 were found to have significant causal associations with GBM. Notably, UDIPs such as T1-33 (OR = 1.007, 95% CI = 1.001 to 1.012, P = .022), T1-34 (OR = 1.012, 95% CI = 1.001-1.023, P = .028), and T1-96 (OR = 1.009, 95% CI = 1.001-1.019, P = .046) were found to have a genetic association with GBM. Furthermore, T1-34 and T1-96 were significantly associated with GBM recurrence, with P-values < .0001 and P < .001, respectively. In addition, scPagwas analysis revealed that T1-33, T1-34, and T1-96 are distinctively linked to different GBM subtypes, with T1-33 showing strong associations with the neural progenitor-like subtype (NPC2), T1-34 with mesenchymal (MES2) and neural progenitor (NPC1) cells, and T1-96 with the NPC2 subtype. T1-33, T1-34, and T1-96 hold significant potential for predicting tumor recurrence and aiding in the development of personalized GBM treatment strategies.

PMID:39879386 | DOI:10.1093/bib/bbaf037

PLoS One. 2025 Jan 29;20(1):e0316493. doi: 10.1371/journal.pone.0316493. eCollection 2025.

ABSTRACT

The emergence of Next Generation Sequencing (NGS) technology has catalyzed a paradigm shift in clinical diagnostics and personalized medicine, enabling unprecedented access to high-throughput microbiome data. However, the inherent high dimensionality, noise, and variability of microbiome data present substantial obstacles to conventional statistical methods and machine learning techniques. Even the promising deep learning (DL) methods are not immune to these challenges. This paper introduces a novel feature engineering method that circumvents these limitations by amalgamating two feature sets derived from input data to generate a new dataset, which is then subjected to feature selection. This innovative approach markedly enhances the Area Under the Curve (AUC) performance of the Deep Neural Network (DNN) algorithm in colorectal cancer (CRC) detection using gut microbiome data, elevating it from 0.800 to 0.923. The proposed method constitutes a significant advancement in the field, providing a robust solution to the intricacies of microbiome data analysis and amplifying the potential of DL methods in disease detection.

PMID:39879257 | DOI:10.1371/journal.pone.0316493

Ann Am Thorac Soc. 2025 Jan 29. doi: 10.1513/AnnalsATS.202409-906OC. Online ahead of print.

ABSTRACT

RATIONALE: Tobacco smoking is a well-established risk factor for idiopathic pulmonary fibrosis (IPF), yet the influence of early-life tobacco exposure on future IPF risk remains poorly understood.

OBJECTIVES: To test the hypothesis that early-life tobacco exposure may elevate the risk of developing IPF, with this effect potentially modified by genetic susceptibility to IPF and mediated through accelerated biological aging.

METHODS: Using data from over 430,000 participants in the UK Biobank, we performed a prospective cohort study to examine the associations of maternal smoking around birth and age of smoking initiation with IPF risk. We evaluated the combined effects and interactions between early-life tobacco exposure and genetic susceptibility to IPF, quantified using polygenic risk scores. We assessed biological aging, as measured by telomere length and phenotypic age, as potential mediators in the associations between early-life tobacco exposure and IPF risk. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs).

RESULTS: Maternal smoking around birth was associated with a higher risk of IPF (HR: 1.26; 95% CI: 1.11-1.43). Compared to never-smokers, individuals who initiated smoking in childhood (HR: 3.65; 95% CI: 3.02-4.41), adolescence (HR: 2.64; 95% CI: 2.28-3.05), and adulthood (HR: 2.09; 95% CI: 1.79-2.44) exhibited increased IPF risk (P for trend < 0.001). An additive interaction was observed between age of smoking initiation and genetic risk for IPF. Individuals with high genetic risk, maternal smoking exposure, and childhood smoking initiation had a 16-fold greater risk of IPF (HR: 16.47; 95% CI: 9.57-28.32), compared to those with low genetic risk and no tobacco exposure. Telomere length and phenotypic age each mediated approximately 10% of the effect of maternal smoking on IPF, with weaker mediation effects observed for later ages of smoking initiation.

CONCLUSION: Early-life tobacco exposure may elevate the risk of IPF, with effect modified by genetic susceptibility and partially mediated through accelerated biological aging.

PMID:39879538 | DOI:10.1513/AnnalsATS.202409-906OC

PLoS One. 2025 Jan 29;20(1):e0316484. doi: 10.1371/journal.pone.0316484. eCollection 2025.

ABSTRACT

BACKGROUND AND AIMS: Predicting progression and prognosis in Interstitial Lung Diseases (ILD), especially Idiopathic Pulmonary Fibrosis (IPF) and Progressive Pulmonary Fibrosis (PPF), remains a challenge. Integrating patient-centered measurements is essential for earlier and safer detection of disease progression. Home monitoring through e-health technologies, such as spirometry and oximetry connected to smartphone applications, holds promise for early detection of ILD progression or acute exacerbations, enabling timely therapeutic interventions.

METHODS: The European ILD Registry Algorithm for Self-Assessment in ILD (eurILDreg ASA-ILD), developed by all eurILDreg principal investigators, includes questionnaires on symptom burden, respiratory infections, and quality of life (EQ5D VAS, K-BILD, LCQ). The algorithm also incorporates spirometry and oxygen saturation measurements, both at rest and during exercise (one-minute sit-to-stand test, 1STST). This ASA-ILD algorithm is integrated into the patientMpower Ltd. smartphone application, used for patient-led monitoring, research, and clinical care since 2016, and available on both Apple and Android platforms.

DISCUSSION: For patient-centered measurements, participants in the multicenter eurILDreg study will receive a patientMpower account, a handheld clinical-grade spirometer (Spirobank Smart, MIR, Italy), and a pulse oximeter (Nonin Medical, Inc. Plymouth, MN, USA), along with usage instructions. Artificial intelligence software (ArtiQ) will analyze spirometry maneuvers in real-time, ensuring compliance with recent ERS/ATS criteria and providing automated feedback. Pulse oximetry is integrated into the exercise testing within the application, following an automated in-app protocol developed with clinician involvement for safety and accuracy. The application will send reminders to participants to complete patient-reported outcome measures (PROMs) according to the study protocol.

CONCLUSION: This study is designed to explore the potential of e-Health technologies, such as home monitoring via spirometry and oximetry, integrated with the eurILDreg ASA-ILD algorithm and patientMpower app, to improve early detection and management of ILD. A pilot trial showed promising adherence to spirometry, indicating that digital health interventions could enhance patient care and outcomes in ILD.

TRIAL REGISTRATION: The ethics committee of the Justus-Liebig-University of Giessen has approved the eurILDreg and this substudy with the protocol reference number 111/08. The research was conducted strictly according to the principles of the Declaration of Helsinki. Patients were included into the registry upon having signed the informed consent. The eurIPFreg and eurIPFbank are listed in ClinicalTrials.gov (NCT02951416). EurILDreg is registered in German Clinical Trials Register, DRKS 00028968.

PMID:39879227 | DOI:10.1371/journal.pone.0316484