Front Cell Infect Microbiol. 2025 May 13;15:1455735. doi: 10.3389/fcimb.2025.1455735. eCollection 2025.

ABSTRACT

BACKGROUND: Probiotics are recognized as beneficial foods, but adverse reactions reported by individuals still exist. This study aims to analysis adverse events (AE) related to probiotics from the FAERS database from the first quarter (Q1) of 2005 to the fourth quarter (Q4) of 2023.

METHODS: The AE data related to probiotic from the 2005 Q1 to the 2023 Q4 were collected. R language was applied to analyze the standardized AE data and three algorithms including the reporting odds ratio (ROR), the proportional reporting ratio (PRR) and the empirical Bayes geometric mean (EBGM) were used to identify AE signals.

RESULTS: In this study, 10,698,312 reports were collected from the FAERS database, of which 74 probiotic-related adverse events were reported. About one third of the reported cases were older than 60 years.36.36% of the reported cases required Hospitalization. A total of 285 preference terms (PTS) and 15 system organ classes (SOC) were identified. In the overall analysis, only 9 PTs and 2 SOCs met significant disproportionality for all three algorithms simultaneously. SOCs included Gastrointestinal disorders (N=97, ROR=5.3, PRR=3.84, EBGM=3.84) and Hepatobiliary disorders (N=9, ROR =3.39, PRR=3.32, EBGM=3.32). PTs included Gastrointestinal pain (ROR=77.76, PRR=76.69, EBGM=76.63), Hypophagia (ROR=24.13, PRR=23.88, EBGM=28.88), and Hepatobiliary disorders (N=97, ROR=5.3, PRR=3.84, EBGM=3.84) and Flatulence (ROR=23.75, PRR=23.28, EBGM=23.27) were the top four highest. Meanwhile, s found new unique adverse signals such as Agitation (ROR=12.48, PRR=12.32, EBGM=12.32) and Anxiety (ROR=4.10, PRR=4.04, EBGM=4.04). Additionally, subgroup analyses were performed to identify AE signals based on gender and age. Metabolism and nutrition disorders (N=6, ROR=3.21, PRR=3.04, EBGM=3.04) and Asthenia (N=3, ROR=5.9, PRR=5.71, EBGM=5.71) were unique AE signal for the male group.

CONCLUSION: Although, the risk of adverse reactions arising from the application of probiotics cannot be ignored. However, However, the results of this FAERS-based study continue to support the overall safety of probiotic preparations. It is necessary to pay attention to the potential influence of factors such as gender and age on the effects and adverse reactions of probiotic application in basic research and clinical application.

PMID:40433664 | PMC:PMC12106448 | DOI:10.3389/fcimb.2025.1455735

J Res Pharm Pract. 2025 Apr 24;13(4):111-118. doi: 10.4103/jrpp.jrpp_2_25. eCollection 2024 Oct-Dec.

ABSTRACT

OBJECTIVE: In severe cases, COVID-19 can lead to a hyperinflammatory state, resulting in devastating outcomes. Immune modulation using steroids or other immune modulators can regulate the intensity of the inflammatory response; however, this theory has not been adequately assessed in practice. The current study aims to investigate the use of corticosteroids alone or in combination with tocilizumab to treat patients with severe COVID-19.

METHODS: This cross-sectional study was conducted on 166 Iranian patients with severe COVID-19 infection at Al-Zahra Hospital, who were treated with the standard treatment for severe COVID-19 infection, as per the 11th version of the Iranian guideline for COVID-19 treatment. Patients were categorized into three treatment groups based on the dose of corticosteroid treatment and tocilizumab therapy: (a) high-dose methylprednisolone (>1 mg/kg) alone, (b) low-dose methylprednisolone (<1 mg/kg) followed by one dose of tocilizumab (8 mg/kg); and (c) high-dose methylprednisolone (>1 mg/kg) followed by one dose of tocilizumab (8 mg/kg). Mortality of patients as our primary outcome, laboratory parameters, length of hospitalization, intensive care unit (ICU) admission requirement, and drug-related adverse events were compared between groups.

FINDINGS: The second group showed significantly better outcomes, including shorter ICU stays, lower C-reactive protein and lactate dehydrogenase levels, and higher oxygen saturation and platelet counts than the other groups. Logistic regression revealed increased risks of mortality, nosocomial infection, and adverse effects, including hepatic and renal dysfunction and gastrointestinal bleeding, in Groups B and C compared with Group A.

CONCLUSION: In all evaluated parameters, a low-dose steroid followed by tocilizumab was superior to a high-dose steroid alone or combined with tocilizumab. Although this combination treatment has been assessed worldwide, few studies have focused on its application in Iranian patients with severe COVID-19.

PMID:40432839 | PMC:PMC12105767 | DOI:10.4103/jrpp.jrpp_2_25

Medicina (Kaunas). 2025 May 17;61(5):911. doi: 10.3390/medicina61050911.

ABSTRACT

Tuberculosis remains a significant public health challenge globally. The emergence of multidrug-resistant Mycobacterium tuberculosis strains presents one of the biggest hurdles in tuberculosis management. Both first- and second-line tuberculosis drugs are associated with common adverse reactions, which can lead to treatment interruptions and decreased adherence. In this article, we review the most commonly used drugs for the treatment of tuberculosis, focusing on the adverse reactions they may cause. We will examine the frequency and timeline of adverse drug reactions involving gastrointestinal, cardiac, neurological, nephrological, and cutaneous systems. Identifying patients at risk of developing those reactions is crucial for healthcare providers to implement monitoring strategies and manage complications effectively. In the review, we present the data about risk factors, management recommendations, and drug discontinuation rates as a result of side effects.

PMID:40428869 | DOI:10.3390/medicina61050911

Bioinformatics. 2025 Jun 2;41(6):btaf319. doi: 10.1093/bioinformatics/btaf319.

ABSTRACT

MOTIVATION: Drug side effects refer to harmful or adverse reactions that occur during drug use, unrelated to the therapeutic purpose. A core issue in drug side effect prediction is determining the frequency of these drug side effects in the population, which can guide patient medication use and drug development. Many computational methods have been developed to predict the frequency of drug side effects as an alternative to clinical trials. However, existing methods typically build regression models on five frequency classes of drug side effects and tend to overfit the training set, leading to boundary handling issues and the risk of overfitting.

RESULTS: To address this problem, we develop a multi-source similarity fusion-based model, named multi-source similarity fusion (MSSF), for predicting five frequency classes of drug side effects. Compared to existing methods, our model utilizes the multi-source feature fusion module and the self-attention mechanism to explore the relationships between drugs and side effects deeply and employs Bayesian variational inference to more accurately predict the frequency classes of drug side effects. The experimental results indicate that MSSF consistently achieves superior performance compared to existing models across multiple evaluation settings, including cross-validation, cold-start experiments, and independent testing. The visual analysis and case studies further demonstrate MSSF's reliable feature extraction capability and promise in predicting the frequency classes of drug side effects.

AVAILABILITY AND IMPLEMENTATION: The source code of MSSF is available on GitHub (https://github.com/dingxlcse/MSSF.git) and archived on Zenodo (DOI: 10.5281/zenodo.15462041).

PMID:40424358 | DOI:10.1093/bioinformatics/btaf319

JAMA Netw Open. 2025 May 1;8(5):e2514278. doi: 10.1001/jamanetworkopen.2025.14278.

ABSTRACT

IMPORTANCE: Available antidepressants provide inadequate therapeutic responses in many patients with major depressive disorder (MDD), highlighting a substantial unmet need.

OBJECTIVE: To evaluate the efficacy and safety of azetukalner, a novel, potent KV7 potassium channel opener, in participants with MDD.

DESIGN, SETTING, AND PARTICIPANTS: X-NOVA was a multicenter, proof-of-concept, phase 2, randomized, double-blind, parallel-group, placebo-controlled clinical trial that evaluated azetukalner in participants (adults aged ≥18 to ≤65 years) with moderate to severe MDD in a current depressive episode. Participants were enrolled between April 2022 and October 2023, and data analysis occurred from January 2023 to January 2024.

INTERVENTION: Participants were randomized (1:1:1) to 10 mg of azetukalner, 20 mg of azetukalner, or placebo orally once daily with food for 6 weeks, with a 4-week follow-up. Concomitant antidepressant medications were not permitted.

MAIN OUTCOMES AND MEASURES: The primary efficacy end point was change in Montgomery-Åsberg Depression Rating Scale (MADRS) score at week 6. Secondary end points included change from baseline at week 6 in the Snaith-Hamilton Pleasure Scale (SHAPS) and Beck Anxiety Inventory. Exploratory end points included change in the Hamilton Depression Rating Scale, 17-Item (HAM-D17) score and change in MADRS at week 1. Frequency and severity of treatment-emergent adverse events (TEAEs) were recorded.

RESULTS: Altogether, 168 participants were randomized (56 to placebo, 56 to 10 mg of azetukalner, and 56 to 20 mg of azetukalner); mean (SD) age was 47.2 (13.6) years, and 111 participants (66.5%) were female. The modified intent-to-treat and safety populations consisted of 164 and 167 participants, respectively. The mean (SE) reduction in MADRS scores from baseline to week 6 was -13.90 (1.41) points with placebo, -15.61 (1.34) points with 10 mg of azetukalner, and -16.94 (1.45) points with 20 mg of azetukalner; the mean (SE) reduction with 20 mg of azetukalner vs placebo was clinically meaningful but not statistically significant (-3.04 points; 95% CI, -7.04 to 0.96 points; P = .14) at week 6, while significant at week 1 (-2.66 points; 95% CI, -5.30 to -0.03 points; P = .047). The mean (SE) reduction in HAM-D17 from baseline to week 6 was significantly greater with 20 mg of azetukalner vs placebo (-13.3 [1.1] vs -10.2 [1.0] points; P = .04). The mean (SE) reduction in SHAPS scores from baseline to week 6 was significantly greater with 20 mg of azetukalner vs placebo (-7.77 [0.87] vs -5.30 [0.85] points; P = .046). Similar rates of discontinuation due to TEAEs were reported across groups.

CONCLUSIONS AND RELEVANCE: In this randomized clinical trial of azetukalner, preliminary findings supported its further clinical development for the treatment of MDD and anhedonia.

TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT05376150.

PMID:40423966 | PMC:PMC12117446 | DOI:10.1001/jamanetworkopen.2025.14278

BMC Med Res Methodol. 2025 May 27;25(1):146. doi: 10.1186/s12874-025-02597-z.

ABSTRACT

BACKGROUND: The widespread use of long-term pharmacological treatments for chronic conditions has led to polypharmacy, raising concerns about adverse effects and interactions. Deprescribing, the discontinuation of drugs with unfavorable benefit-risk ratios, is gaining attention. Studies evaluating the discontinuation of drugs have a broad methodological spectrum. The selection of outcomes poses a particular challenge. This scoping review addresses the methodological challenges of outcome selection in RCTs investigating drug discontinuation.

METHODS: The scoping review includes RCTs that investigated the discontinuation of drugs whose efficacy and/or safety was in doubt. Data on study characteristics, the motivation for evaluating drug discontinuation, the number and type of primary endpoints, and the stated hypotheses were extracted and analyzed.

RESULTS: We included 103 RCTs. Most studies were from Europe and the USA and mainly investigated antipsychotics/antidepressants, immunosuppressants, steroids and antiepileptics. The discontinuation studies were often conducted due to side effects of the treatment and doubts about the benefits of the drug. The primary endpoints reflected either the course of the disease ("justification of treatment") or the disadvantages of the drug ("justification of withdrawal"). Non-inferiority hypotheses were generally prevalent in justification of treatment studies, while superiority hypotheses were more commonly used in justification of withdrawal studies. However, due to methodological and practical challenges this was not always the case.

CONCLUSION: We present a framework to choose outcomes and specify hypotheses for discontinuation studies. With regard to this, both key challenges (justification of treatment and justification of withdrawal) must be met.

PMID:40426033 | DOI:10.1186/s12874-025-02597-z

Sci Data. 2025 May 27;12(1):872. doi: 10.1038/s41597-025-04906-z.

ABSTRACT

Pathology data, primarily consisting of slides and diagnostic reports, inherently contain knowledge that is pivotal for advancing data-driven biomedical research and clinical practice. However, the hidden and fragmented nature of this knowledge across various data modalities not only hinders its computational utilization, but also impedes the effective integration of AI technologies within the domain of pathology. To systematically organize pathology knowledge for its computational use, we propose PathoGraph, a knowledge representation method that describes pathology knowledge in a graph-based format. PathoGraph can represent: (1) pathological entities' types and morphological features; (2) the composition, spatial arrangements, and dynamic behaviors associated with pathological phenotypes; and (3) the differential diagnostic approaches used by pathologists. By applying PathoGraph to neoplastic diseases, we illustrate its ability to comprehensively and structurally capture multi-scale disease characteristics alongside pathologists' expertise. Furthermore, we validate its computational utility by demonstrating the feasibility of large-scale automated PathoGraph construction, showing performance improvements in downstream deep learning tasks, and presenting two illustrative use cases that highlight its clinical potential. We believe PathoGraph opens new avenues for AI-driven advances in the field of pathology.

PMID:40425649 | DOI:10.1038/s41597-025-04906-z

Thorac Cancer. 2025 May;16(10):e70048. doi: 10.1111/1759-7714.70048.

ABSTRACT

BACKGROUND: Lymph node metastasis (LNM) poses a considerable threat to survival in lung adenocarcinoma. Currently, minor resection is the recommended surgical approach for small-diameter lung cancer. The accurate preoperative identification of LNM in patients with small-diameter lung cancer is important for improving patient survival and outcomes.

METHODS: A total of 1740 patients with clinical early-stage lung adenocarcinoma who underwent surgical resection were enrolled in this study. The Lasso model was used to screen clinical and imaging features, and multivariate logistic regression analysis was used to analyze the relevant diagnostic factors to establish a diagnostic model for predicting LNM. Receiver operating characteristic (ROC) curve analysis, decision curve analysis (DCA) and calibration curve analysis were used to verify the clinical efficacy of the model, which was further validated with an internal validation set.

RESULTS: The proportion of solid components (PSC), sphericity, nodule margin, entropy, and edge blur were identified as diagnostic factors that were strongly correlated with LNM in lung adenocarcinoma patients. The area under the ROC curve (AUC) in the internal training set was 0.91. Decision curve analysis revealed that the model could achieve greater benefits for patients. The calibration curve was used to further verify the applicability of the prediction model.

CONCLUSIONS: Patients with early-stage lung adenocarcinoma with LNM can be identified by typical imaging features. The diagnostic model can help to optimize surgical planning among thoracic surgeons.

PMID:40425526 | DOI:10.1111/1759-7714.70048

J Neural Eng. 2025 May 27. doi: 10.1088/1741-2552/addd4a. Online ahead of print.

ABSTRACT

Objective: Magnetoencephalography (MEG) is a cutting-edge neuroimaging technique that measures the intricate brain dynamics underlying cognitive processes with an unparalleled combination of high temporal and spatial precision. While MEG data analytics have traditionally relied on advanced signal processing and mathematical and statistical tools, the recent surge in Artificial Intelligence (AI) has led to the growing use of Machine Learning (ML) methods for MEG data classification. An emerging trend in this field is the use of Artificial Neural Networks (ANNs) to address various MEG-related tasks. This review aims to provide a comprehensive overview of the state of the art in this area.Approach: This topical review included studies that applied ANNs to MEG data. Studies were sourced from PubMed, Google Scholar, arXiv, and bioRxiv using targeted search queries. The included studies were categorized into three groups: Classification, Modeling, and Other. Key findings and trends were summarized to provide a comprehensive assessment of the field.Main Results:The review identified 119 relevant studies, with 69 focused on Classification, 16 on Modeling, and 34 in the Other category. Classification studies addressed tasks such as brain decoding, clinical diagnostics, and BCI implementations, often achieving high predictive accuracy. Modeling studies explored the alignment between ANN activations and brain processes, offering insights into the neural representations captured by these networks. The Other category demonstrated innovative uses of ANNs for artifact correction, preprocessing, and neural source localization.Significance: By establishing a detailed portrait of the current state of the field, this review highlights the strengths and current limitations of ANNs in MEG research. It also provides practical recommendations for future work, offering a helpful reference for seasoned researchers and newcomers interested in using ANNs to explore the complex dynamics of the human brain with MEG.

PMID:40425030 | DOI:10.1088/1741-2552/addd4a

Sci Rep. 2025 May 27;15(1):18533. doi: 10.1038/s41598-025-02999-5.

ABSTRACT

Red radish microgreens (RRM) have gained considerable attention for their promising therapeutic potential. However, the molecular mechanisms underlying their bioactivity remain inadequately characterized. This study explores the anti-inflammatory, antioxidant, and anticancer properties of RRM extract using in silico and in vivo Drosophila model analyses. The metabolite profile of the RRM extract was characterized using comprehensive metabolomics techniques, including Gas Chromatography-Mass Spectrometry (GC-MS) and Liquid Chromatography High-Resolution Mass Spectrometry (LC-HRMS). Furthermore, in silico analysis utilizing network pharmacology identified target proteins of RRM compounds associated with cancer, inflammation, and oxidative stress. Concurrently, in vivo experiments with Drosophila melanogaster PGRP-LBΔ (Dm PGRP-LBΔ) larvae was conducted to assess the extract's impact on immune and oxidative stress pathways. In silico analysis revealed that RRM compounds interacted with key proteins (AKT1, ESR1, MAPK1, SRC, TP53), modulating pathways related to cancer, inflammation, and oxidative stress. Molecular dynamics simulations reinforced the docking results by confirming robust binding of kaempferitrin to AKT1. In vivo studies showed that RRM extract suppressed immune-related genes (dptA, totA) through the NFκB and JAK-STAT pathways, reduced ROS levels, and selectively regulated antioxidant gene expression by enhancing sod1 while decreasing sod2 and cat. These results suggest RRM extract as a functional food for managing oxidative stress, inflammation, and cancer. Further research in higher organisms and clinical settings is needed.

PMID:40425671 | DOI:10.1038/s41598-025-02999-5

Infection. 2025 May 27. doi: 10.1007/s15010-025-02559-z. Online ahead of print.

ABSTRACT

Hemophagocytic lymphohistiocytosis (HLH) is an orphan disease characterized by excessive inflammation and poor outcome. We sought to further characterize clinical features, courses, and risk factors of secondary HLH (sHLH) triggered by infection (iHLH). 28 (43.1%) of 65 adult sHLH cases treated at our hospital from 2012-2024 were infection-associated. iHLH patients were mostly male (71.4%). Infectious agents most frequently detected were EBV (57.1%) and leishmania (14.3%). The median time to diagnosis was 13 [6.0;24.8] days. iHLH patients had a mortality rate of 39.3% (median follow-up time: 735 [336;1140] days), worse survival than patients with autoimmune-triggered (hazard ratio: 3.33 (1.01-11.10), p = 0.049), and better survival than patients with paraneoplastic HLH (hazard ratio: 0.19 (0.10-0.84), p = 0.002). Elevated levels of soluble interleukin-2 receptor (sIL2R; > 6,000 I/U), low thrombocyte counts (< 40 G/l), and a history of malignant disease were associated with adverse outcomes. Protracted time to diagnosis was associated with severe disease courses and with leishmaniosis. Further, sIL2R levels correlated positively with prolonged aPTT and thrombocytopenia, and hypertriglyceridemia with elevated INRs. Patients with an elevated sIL2R:ferritin ratio were more likely to have a history of malignant comorbidities. Taken together, sIL2R, thrombocytopenia, and a history of malignant disease are important prognostic factors of iHLH. Patients with high sIL2R levels or hypertriglyceridemia may be at higher risk of bleeding, and patients with elevated sIL2R:ferritin ratios should be assessed for possible malignant comorbidities. Lastly, increased awareness of the disease and newly emerging pathogens (i.e. leishmania) may shorten the time to diagnosis, and thus reduce severe courses of iHLH.

PMID:40425997 | DOI:10.1007/s15010-025-02559-z

J Cyst Fibros. 2025 May 26:S1569-1993(25)00111-0. doi: 10.1016/j.jcf.2025.03.015. Online ahead of print.

ABSTRACT

BACKGROUND: Lenabasum is a cannabinoid receptor 2 (CB2) agonist under development for cystic fibrosis (CF), targeting inflammation. We evaluated the efficacy and safety of lenabasum in people with CF (pwCF).

METHODS: We conducted a global, 28-week, randomized, double-blind, placebo-controlled Phase 2b trial. PwCF were ≥12 years old with 2-3 pulmonary exacerbations (PEx) treated with intravenous (IV) antibiotics (or 1 PEx treated with IV and ≥1 PEx treated with oral antibiotics) in the past year. Subjects were randomized 2:1:2 to lenabasum 20 mg BID, lenabasum 5 mg BID, or placebo BID. Primary endpoint was rate of PEx, comparing lenabasum 20 mg BID to placebo.

RESULTS: Among 447 subjects from 21 countries, mean age was 26.9 (10.3 SD) years, 53.6% were female, 45.2% homozygous for F508del, and 24.9% received CFTR modulators. Highest ppFEV1 in the previous year was 69.2% with the majority having 1-2 PEx treated with IV antibiotics (2-7 PEx treated with either IV or oral antibiotics). PEx incidence over 28 weeks was 0.84 for placebo, 0.75 for lenabasum 5 mg BID, and 0.91 for lenabasum 20 mg BID; rates were not lower relative to placebo in the 5 mg (incidence rate ratio (IRR)=0.89, 95% CI 0.66 to 1.19, p = 0.44) or the 20 mg group (IRR 1.08, 95% CI 0.86 to 1.37, p = 0.51). PEx occurred less frequently in participants from Eastern Europe, but there was no evidence of regional variation in treatment efficacy. Lenabasum was well tolerated, without safety signals.

CONCLUSION: Lenabasum did not improve key clinical outcomes in this Phase 2b study in pwCF.

PMID:40425421 | DOI:10.1016/j.jcf.2025.03.015

Respir Med. 2025 May 25:108179. doi: 10.1016/j.rmed.2025.108179. Online ahead of print.

ABSTRACT

Bronchiectasis is a chronic inflammatory airway disease characterized by a self-perpetuating vortex of impaired mucociliary clearance, persistent infection, and progressive structural lung damage. While inflammation is central to disease activity and progression, targeted anti-inflammatory treatments have been limited. Understanding the different types of inflammation involved and their significant overlap is essential for effective management. This review explores key inflammation patterns, biomarkers, and available treatments across the spectrum of inflammation in bronchiectasis, with a particular focus on non-cystic fibrosis bronchiectasis in adults. Neutrophilic inflammation remains the hallmark of bronchiectasis, with promising reversible dipeptidyl peptidase-1 inhibitors reducing the activation of neutrophil serine proteases during neutrophil maturation. Eosinophilic inflammation has also gained attention, with evidence indicating that patients with this endotype may benefit from glucocorticoids and biologic therapies targeting type 2 inflammation. Additional inflammatory mechanisms discussed here include impaired epithelial function and mucociliary abnormalities, immune dysregulation, and airway inflammation triggered by infections, environmental irritants, and autoimmune conditions. Written for general clinicians, this review simplifies complex concepts, underscores key aspects of diagnostic evaluation, and discusses both conventional and emerging treatments for bronchiectasis, providing practical insights for improved personalized patient care.

PMID:40425105 | DOI:10.1016/j.rmed.2025.108179

BMC Med Imaging. 2025 May 27;25(1):194. doi: 10.1186/s12880-025-01712-2.

ABSTRACT

BACKGROUND: Pineal region tumors (PRTs) are rare but deep-seated brain tumors, and complete surgical resection is crucial for effective tumor treatment. The choice of surgical approach is often challenging due to the low incidence and deep location. This study aims to combine machine learning and deep learning algorithms with pre-operative MRI images to build a model for PRTs surgical approaches recommendation, striving to model clinical experience for practical reference and education.

METHODS: This study was a retrospective study which enrolled a total of 173 patients diagnosed with PRTs radiologically from our hospital. Three traditional surgical approaches of were recorded for prediction label. Clinical and VASARI related radiological information were selected for machine learning prediction model construction. And MRI images from axial, sagittal and coronal views of orientation were also used for deep learning craniotomy approach prediction model establishment and evaluation.

RESULTS: 5 machine learning methods were applied to construct the predictive classifiers with the clinical and VASARI features and all methods could achieve area under the ROC (Receiver operating characteristic) curve (AUC) values over than 0.7. And also, 3 deep learning algorithms (ResNet-50, EfficientNetV2-m and ViT) were applied based on MRI images from different orientations. EfficientNetV2-m achieved the highest AUC value of 0.89, demonstrating a significant high performance of prediction. And class activation mapping was used to reveal that the tumor itself and its surrounding relations are crucial areas for model decision-making.

CONCLUSION: In our study, we used machine learning and deep learning to construct surgical approach recommendation models. Deep learning could achieve high performance of prediction and provide efficient and personalized decision support tools for PRTs surgical approach.

CLINICAL TRIAL NUMBER: Not applicable.

PMID:40426149 | DOI:10.1186/s12880-025-01712-2

BMC Cancer. 2025 May 27;25(1):953. doi: 10.1186/s12885-025-14354-y.

ABSTRACT

BACKGROUND: Diffusion-weighted imaging with higher b-value improves detection rate for prostate cancer lesions. However, obtaining high b-value DWI requires more advanced hardware and software configuration. Here we use a novel deep learning network, NAFNet, to generate a deep learning reconstructed (DLR1500) images from 800 b-value to mimic 1500 b-value images, and to evaluate its performance and lesion detection improvements based on whole-slide images (WSI).

METHODS: We enrolled 303 prostate cancer patients with both 800 and 1500 b-values from Fudan University Shanghai Cancer Centre between 2017 and 2020. We assigned these patients to the training and validation set in a 2:1 ratio. The testing set included 36 prostate cancer patients from an independent institute who had only preoperative DWI at 800 b-value. Two senior radiology doctors and two junior radiology doctors read and delineated cancer lesions on DLR1500, original 800 and 1500 b-values DWI images. WSI were used as the ground truth to assess the lesion detection improvement of DLR1500 images in the testing set.

RESULTS: After training and generating, within junior radiology doctors, the diagnostic AUC based on DLR1500 images is not inferior to that based on 1500 b-value images (0.832 (0.788-0.876) vs. 0.821 (0.747-0.899), P = 0.824). The same phenomenon is also observed in senior radiology doctors. Furthermore, in the testing set, DLR1500 images could significantly enhance junior radiology doctors' diagnostic performance than 800 b-value images (0.848 (0.758-0.938) vs. 0.752 (0.661-0.843), P = 0.043).

CONCLUSIONS: DLR1500 DWIs were comparable in quality to original 1500 b-value images within both junior and senior radiology doctors. NAFNet based DWI enhancement can significantly improve the image quality of 800 b-value DWI, and therefore promote the accuracy of prostate cancer lesion detection for junior radiology doctors.

PMID:40426115 | DOI:10.1186/s12885-025-14354-y

J Imaging Inform Med. 2025 May 27. doi: 10.1007/s10278-025-01542-2. Online ahead of print.

ABSTRACT

Accurate assessment of computed tomography (CT) image quality is crucial for ensuring diagnostic accuracy, optimizing imaging protocols, and preventing excessive radiation exposure. In clinical settings, where high-quality reference images are often unavailable, developing no-reference image quality assessment (NR-IQA) methods is essential. Recently, CT-NR-IQA methods using deep learning have been widely studied; however, significant challenges remain in handling multiple degradation factors and accurately reflecting real-world degradations. To address these issues, we propose a novel CT-NR-IQA method. Our approach utilizes a dataset that combines two degradation factors (noise and blur) to train convolutional neural network (CNN) models capable of handling multiple degradation factors. Additionally, we leveraged RadImageNet pre-trained models (ResNet50, DenseNet121, InceptionV3, and InceptionResNetV2), allowing the models to learn deep features from large-scale real clinical images, thus enhancing adaptability to real-world degradations without relying on artificially degraded images. The models' performances were evaluated by measuring the correlation between the subjective scores and predicted image quality scores for both artificially degraded and real clinical image datasets. The results demonstrated positive correlations between the subjective and predicted scores for both datasets. In particular, ResNet50 showed the best performance, with a correlation coefficient of 0.910 for the artificially degraded images and 0.831 for the real clinical images. These findings indicate that the proposed method could serve as a potential surrogate for subjective assessment in CT-NR-IQA.

PMID:40425960 | DOI:10.1007/s10278-025-01542-2

J Imaging Inform Med. 2025 May 27. doi: 10.1007/s10278-025-01557-9. Online ahead of print.

ABSTRACT

Diffuse midline glioma (DMG) H3 K27M-altered is a rare pediatric brainstem cancer with poor prognosis. To advance the development of predictive models to gain a deeper understanding of DMG, there is a crucial need for seamlessly integrating automatic and highly accurate tumor segmentation techniques. There is only one method that tries to solve this task in this cancer; for that reason, this study develops a modified CNN-based 3D-Unet tool to automatically segment DMG in an accurate way in magnetic resonance (MR) images. The dataset consisted of 52 DMG patients and 70 images, each with T1W and T2W or FLAIR images. Three different datasets were created: T1W images, T2W or FLAIR images, and a combined set of T1W and T2W/FLAIR images. Denoising, bias field correction, spatial resampling, and normalization were applied as preprocessing steps to the MR images. Patching techniques were also used to enlarge the dataset size. For tumor segmentation, a 3D U-Net architecture with residual blocks was used. The best results were obtained for the dataset composed of all T1W and T2W/FLAIR images, reaching an average Dice Similarity Coefficient (DSC) of 0.883 on the test dataset. These results are comparable to other brain tumor segmentation models and to state-of-the-art results in DMG segmentation using fewer sequences. Our results demonstrate the effectiveness of the proposed 3D U-Net architecture for DMG tumor segmentation. This advancement holds potential for enhancing the precision of diagnostic and predictive models in the context of this challenging pediatric cancer.

PMID:40425959 | DOI:10.1007/s10278-025-01557-9

J Imaging Inform Med. 2025 May 27. doi: 10.1007/s10278-025-01549-9. Online ahead of print.

ABSTRACT

This study aimed to develop a fully automated semantic placenta segmentation model that integrates the U-Net and SegNeXt architectures through ensemble learning. A total of 218 pregnant women with suspected placental abnormalities who underwent magnetic resonance imaging (MRI) were enrolled, yielding 1090 annotated images for developing a deep learning model for placental segmentation. The images were standardized and divided into training and test sets. The performance of Placental Segmentation Network (PlaNet-S), which integrates U-Net and SegNeXt within an ensemble framework, was assessed using Intersection over Union (IoU) and counting connected components (CCC) against the U-Net, U-Net + + , and DS-transUNet. PlaNet-S had significantly higher IoU (0.78, SD = 0.10) than that of U-Net (0.73, SD = 0.13) (p < 0.005) and DS-transUNet (0.64, SD = 0.16) (p < 0.005), while the difference with U-Net + + (0.77, SD = 0.12) was not statistically significant. The CCC for PlaNet-S was significantly higher than that for U-Net (p < 0.005), U-Net + + (p < 0.005), and DS-transUNet (p < 0.005), matching the ground truth in 86.0%, 56.7%, 67.9%, and 20.9% of the cases, respectively. PlaNet-S achieved higher IoU than U-Net and DS-transUNet, and comparable IoU to U-Net + + . Moreover, PlaNet-S significantly outperformed all three models in CCC, indicating better agreement with the ground truth. This model addresses the challenges of time-consuming physician-assisted manual segmentation and offers the potential for diverse applications in placental imaging analyses.

PMID:40425958 | DOI:10.1007/s10278-025-01549-9

Aesthetic Plast Surg. 2025 May 27. doi: 10.1007/s00266-025-04924-7. Online ahead of print.

ABSTRACT

BACKGROUND: The biomechanics of frontalis muscle contraction and its interaction with skin remain contentious, particularly the debated bidirectional movement theory. This study introduces the quantum elasticity and resistance gradient (QERG) model to explain observed skin dynamics during frontalis contraction using elastic resistance principles.

METHODS: An AI-driven biomechanical model incorporating deep learning frameworks (TensorFlow, PyTorch) was developed to simulate skin deformation and muscle forces during frontalis contraction. The model was trained using 3D facial scans from a diverse cohort of 600 subjects, representing various ethnicities, genders, and ages. Resistance gradients and wrinkle formation were calculated using finite element analysis, and machine learning (random forest, deep neural networks) was employed to predict skin behaviour.

RESULTS: Cranial displacement averaged 6.9 mm across all subjects, with younger individuals (18-30 years) showing higher displacement than older individuals (50-65 years). Ethnic differences in displacement and wrinkle formation were observed, with Caucasians exhibiting greater displacement (7.3 mm) compared to African Americans and Asians (6.0 mm and 5.8 mm). The QERG model predicted skin folding at an average threshold of 41.2 mm above the eyebrows, with variations linked to ethnicity, age, and gender. AI models achieved high accuracy (R2 = 0.96), validating the model's predictive power.

CONCLUSION: The QERG model confirms that frontalis muscle contraction is unidirectional, with skin folding attributed to elastic resistance rather than opposing forces. These findings challenge previous theories of bidirectional contraction and have implications for aesthetic treatments.

LEVEL OF EVIDENCE III: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

PMID:40425886 | DOI:10.1007/s00266-025-04924-7