Inflammopharmacology. 2025 May 8. doi: 10.1007/s10787-025-01766-2. Online ahead of print.

ABSTRACT

Ototoxicity, the property of certain drugs to cause hearing loss, is a significant concern in medical treatments, particularly with the use of chemotherapeutic agents like cisplatin and aminoglycosides. These drugs can lead to permanent sensorineural hearing loss (SNHL), affecting a substantial proportion of patients. Existing strategies to alleviate these side effects are limited, prompting interest in natural products as potential protective agents. Natural products are being investigated for their ability to counteract these mechanisms through anti-inflammatory and antioxidant properties. The review seeks to highlight the potential of these natural products as complementary therapies to conventional ototoxic medications, emphasizing their protective roles, which are involved in cochlear cellular damage and programmed cell death. Further research is essential to establish standardized protocols for their use and to ensure their integration into clinical practice as effective therapeutic options.

PMID:40338449 | DOI:10.1007/s10787-025-01766-2

Cochrane Database Syst Rev. 2025 May 8;5:CD014585. doi: 10.1002/14651858.CD014585.pub2.

ABSTRACT

RATIONALE: Non-steroidal anti-inflammatory drugs (NSAIDs) are among the most frequently prescribed medicines, commonly used to mitigate pain, inflammation, and cardiovascular prevention, among others. Chronic NSAID consumption increases the risk of acute renal failure, stroke, myocardial infarction, and gastrointestinal toxicity, ranging from mild dyspepsia to serious ulcer complications such as bleeding, obstruction, and perforation. Proton pump inhibitors (PPIs) may exert a gastroprotective effect from NSAID gastroduodenal injury by reducing gastric acid secretion.

OBJECTIVES: To assess the effects of proton pump inhibitors on the prevention of dyspepsia and ulcers in people with chronic consumption of non-steroidal anti-inflammatory drugs.

SEARCH METHODS: We searched CENTRAL, MEDLINE (Ovid), Embase (Ovid), and two trial registers up to 23 October 2023, as well as reference checking, citation searching, and contact with study authors to identify additional studies.

ELIGIBILITY CRITERIA: We included randomised controlled trials (RCTs) and cluster-RCTs comparing PPIs taken orally versus placebo, histamine 2-receptor antagonists, misoprostol, or sucralfate in adults and children with chronic consumption of NSAIDs for four weeks or longer.

OUTCOMES: Our outcomes were global symptoms of dyspepsia, incident ulcer, adverse events, ulcer complications, and quality of life.

RISK OF BIAS: We used the Cochrane RoB 2 tool for RCTs and the tool extension for cluster-RCTs.

SYNTHESIS METHODS: We conducted meta-analyses using random-effects models to calculate risk ratios (RR) and 95% confidence intervals (CI) for dichotomous outcomes and mean differences (MD) and 95% CIs for continuous outcomes. Due to statistical heterogeneity, we conducted meta-analyses for all but two outcomes. We summarised the certainty of evidence according to GRADE methods.

INCLUDED STUDIES: We included 12 studies with 8760 participants. All studies were conducted in an outpatient setting in Africa, Asia, Europe, North America, Central America, South America, and Australia. They were published between 1996 and 2014. All studies measured outcomes in the short term (up to 12 months).

SYNTHESIS OF RESULTS: PPI versus placebo PPIs may have little to no effect on global symptoms of dyspepsia assessed as a dichotomous outcome, but the evidence is very uncertain (meta-analysis was not possible due to high and unexplained statistical heterogeneity and point estimates of RR ranged from 0.36 to 1.13; 8 studies; 4944 participants; very low-certainty evidence). PPIs probably result in a slight reduction in global symptoms of dyspepsia assessed as a continuous outcome (MD -0.56, 95% CI -0.74 to -0.38; 2 studies, 1149 participants; moderate-certainty evidence). PPIs probably result in a reduction in incident ulcers compared to placebo (RR 0.29, 95% CI 0.23 to 0.36; 11 studies, 7022 participants; moderate-certainty evidence). PPIs may have few or no adverse events, but the evidence is very uncertain (meta-analysis was not possible due to high and unexplained statistical heterogeneity and point estimates of RR ranged from 0.67 to 6.35; 12 studies, 7530 participants; very low-certainty evidence). PPIs may reduce ulcer complications compared with placebo (RR 0.33, 95% CI 0.10 to 1.07; P = 0.30; I2 = 18%; 5 studies, 4394 participants; low-certainty evidence). PPIs probably result in a slight increase in quality of life (MD 0.39, 95% CI 0.23 to 0.55; 2 studies, 1149 participants; moderate-certainty evidence). PPI versus histamine 2-receptor antagonists PPIs may increase incident ulcers (RR 2.00, 95% CI 0.21 to 19.44; 1 study, 26 participants; low-certainty evidence). The included study did not report data on global symptoms of dyspepsia, adverse events, ulcer complications, or quality of life. PPI versus misoprostol PPIs may increase incident ulcers (RR 2.32, 95% CI 1.25 to 4.30; 1 study, 402 participants; very low-certainty evidence) and may have fewer adverse events (RR 0.38, 0.25 to 0.57; 1 study, 402 participants; very low-certainty evidence), but the evidence is very uncertain. The included study did not report data on global symptoms of dyspepsia, ulcer complications, or quality of life. No studies compared PPI against sucralfate. Most included studies were at overall high risk of bias or overall some concerns of risk of bias. Imprecision in the effect estimates was also a concern.

AUTHORS' CONCLUSIONS: Compared with placebo, PPIs may have no effect on the presence of global symptoms of dyspepsia and probably result in a slight reduction in global symptoms of dyspepsia scales. PPI probably reduces incident ulcers and may have little to no effect on adverse events. PPIs may reduce ulcer complications and probably slightly increase quality of life. Compared with histamine 2-receptor antagonists, PPIs may increase incident ulcers. The evidence for this comparison came from only one study. Compared with misoprostol, PPIs may increase incident ulcers and may reduce adverse events, but the evidence is very uncertain. The evidence for this comparison came from only one study. The certainty of the evidence for most outcomes and comparisons was low or very low, except for global symptoms of dyspepsia measured as a continuous outcome, incident ulcer, and quality of life in the comparison of PPI versus placebo. Further research is needed to assess the effect of PPIs compared to other active treatments such as sucralfate, misoprostol, or histamine 2-receptor antagonists. Well-designed and reported studies focussing on patient-important outcomes and addressing the methodological limitations found in the present included studies would be informative. These could include different baseline ulcer risks, ages, and types of NSAIDs. Long-term follow-up would be beneficial.

FUNDING: This Cochrane review had no dedicated funding.

REGISTRATION: Protocol (2022): doi.org/10.1002/14651858.CD014585.

PMID:40337979 | DOI:10.1002/14651858.CD014585.pub2

Curr Drug Saf. 2025 May 7. doi: 10.2174/0115748863367086250420011411. Online ahead of print.

ABSTRACT

BACKGROUND: Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are widely accepted for managing Type 2 diabetes mellitus. However, numerous drug-related problems (DRPs) have recently been reported about GLP-1 RAs.

OBJECTIVES: The present descriptive study aimed to compile and report the DRPs of various GLP-1 RAs.

METHODS: The DRPs reported for all the GLP-1 RAs, including exenatide, lixisenatide, liraglutide, dulaglutide, semaglutide, and tirzepatide, were extracted from the category of injury, poisoning, and procedural complications of VigiAccess. The Pharmaceutical Care Network Europe Association (PCNE) classification for drug-related problems (version 9.1) was used to categorize the DRPs into patient-related, healthcare practice-related, and patient- or healthcare practice- related.

RESULTS: Overall, 315952 potential side effects (PSEs) were reported regarding GLP-1 RAs in VigiAccess under injury, poisoning, and procedural complications. Out of 315952 PSE reports, 83210 were DRPs of GLP-1 RAs. Most of them belong to Dulaglutide (23861; 28.68%), followed by tirzepatide (23274; 27.97%), exenatide (18449; 22.17%), semaglutide (11790; 14.97%), and liraglutide (5767; 6.93%). Among the patient-related DRPs, incorrect dose administered (17797; 42.42%), and most of the reports documented for tirzepatide (9993; 23.82%), dulaglutide (4581; 10.92%), and exenatide (2557; 6.10%); however, semaglutide (414; 0.99%), and liraglutide (249; 0.59%), have minor reports documented. Off-label use (13600), most of which were from tirzepatide (4945; 17.59%), followed by semaglutide (4176; 14.85%), liraglutide (1853; 6.59%), exenatide (1530; 5.44%), and dulaglutide (1087; 3.87%).

CONCLUSION: Qualified healthcare practitioners must educate the patients administering the GLP- 1 RAs to minimize preventable DRPs. Also, careful and frequent monitoring of GLP-1 RAs improves therapeutic outcomes by ruling out DRPs. Healthcare practitioners should comply with approved therapeutic guidelines to enhance the quality of GLP-1 RAs treatment.

PMID:40337971 | DOI:10.2174/0115748863367086250420011411

Br J Occup Ther. 2023 Apr;86(4):257-264. doi: 10.1177/03080226221135366. Epub 2023 Mar 21.

ABSTRACT

INTRODUCTION: Medication management is an essential instrumental activity of daily living for older adults; however, 40-70% of older adults fail to take their medications correctly. Addressing medication management falls under the scope of occupational therapy, but there is a lack of evidence supporting occupational therapy interventions to improving medication management. This study's primary aims are to examine the feasibility, acceptability, and preliminary efficacy of a Tailored Intervention for Medication Management delivered by occupational therapists to improve medication management.

METHOD/DESIGN: Single-blind, parallel-group randomized controlled equivalency trial, with two phases. Thirty community-dwelling older adults will be enrolled in this study. In Phase 1, participants in the treatment group will receive Tailored Intervention for Medication Management delivered remotely; those in the waitlist control will receive attention visits. In Phase 2, waitlist control participants will receive Tailored Intervention for Medication Management in person. The primary outcomes are feasibility and acceptability; secondary outcomes include preliminary efficacy of the intervention delivered by an occupational therapist remotely and in person. Additionally, the remote and in-person delivery methods will be compared to each other for equivalency.

DISCUSSION: Inability to manage medication and inappropriate polypharmacy are significant and prevalent problems that must be addressed so older adults can safely perform this essential instrumental activity of daily living.

TRIAL REGISTRATION: ClinicalTrials.gov identifier NCT04717297.

PMID:40337485 | PMC:PMC12033853 | DOI:10.1177/03080226221135366

ERJ Open Res. 2025 May 6;11(3):00623-2024. doi: 10.1183/23120541.00623-2024. eCollection 2025 May.

ABSTRACT

BACKGROUND: Lung transplantation is a highly dynamic segment of solid organ transplantation in which gender plays a central role. Our objective was to investigate the causes of outcome differences between women and men all along the lung transplantation pathway.

METHODS: We used data from the French COhort in Lung Transplantation (COLT) study (12 participating lung transplantation centres). Analyses were performed in three phases: baseline clinical characteristics, peri-transplantation period and post-transplantation follow-up.

RESULTS: Overall, 1710 participants (802 women and 908 men) were included in this study. Women were less likely than men to undergo transplantation (91.6% versus 95.6%; p=0.001) and waited longer before transplantation (115 versus 73 days; p<0.001). Female gender and pre-transplantation class I anti-human leukocyte antigen antibodies were identified as independent factors associated with longer waiting time duration. Female transplant recipients commonly received lungs from height- and sex-matched donors, despite higher female waiting list mortality and a higher proportion of male donors. Importantly, women with oversized lung transplantation (defined by predicted total lung capacity (pTLC) ratio and weight mismatch) did not have worse survival. The overall post-transplantation survival of female recipients was significantly higher than that of male recipients (65.6% versus 57.3%; p<0.001), although the prevalence of specific major lung transplantation outcomes did not differ according to gender.

CONCLUSION: Women waited longer and were less likely to undergo transplantation. Women transplanted with an oversized lung did not have worse survival after transplantation, suggesting that size matching criteria based on pTLC ratio and weight mismatch may be less stringent in this context.

PMID:40337341 | PMC:PMC12053738 | DOI:10.1183/23120541.00623-2024

Thorac Res Pract. 2025 May 8. doi: 10.4274/ThoracResPract.2025.2025-1-1. Online ahead of print.

ABSTRACT

The objective of the study was to assess and compare the efficacy of elexacaftor/tezacaftor/ivacaftor (ELX/TEZ/IVA) treatment with TEZ/IVA treatment in individuals diagnosed with cystic fibrosis (CF) and carrying the F508del allele. An extensive search of relevant literature was conducted using online resources, namely, PubMed, ScienceDirect, and Google Scholar. The initial search identified 248 articles, and after a careful examination of the full text of 18 articles, 7 met the inclusion and exclusion criteria. These selected reports were then thoroughly examined to perform a comparative analysis of the effectiveness of TEZ/IVA versus ELX/TEZ/IVA in CF patients with the F508del allele. The quality of the selected reports was evaluated using the Cochrane risk-of-bias tool for randomized studies, known as RoB 2. ELX/TEZ/IVA has shown significant improvements in key indicators of CF treatment. It has demonstrated a significant increase in forced expiratory volume in one second levels, indicating improved respiratory capacity and airflow. Additionally, ELX/TEZ/IVA successfully reduced sweat chloride levels and positively impacted Cystic Fibrosis Questionnaire-Revised Respiratory Domain scores, reflecting enhanced respiratory function and improved quality of life for patients. Overall, the study concluded that ELX/TEZ/IVA provided a clinically robust benefit compared to TEZ/IVA alone while maintaining a favourable safety profile.

PMID:40336353 | DOI:10.4274/ThoracResPract.2025.2025-1-1

J Exp Orthop. 2025 May 7;12(2):e70259. doi: 10.1002/jeo2.70259. eCollection 2025 Apr.

ABSTRACT

PURPOSE: While several artificial intelligence (AI)-assisted medical imaging applications are reported in the recent orthopaedic literature, comparison of the clinical efficacy and utility of these applications is currently lacking. The aim of this systematic review is to evaluate the effectiveness and reliability of AI applications in orthopaedic imaging, focusing on their impact on diagnostic accuracy, image segmentation and operational efficiency across various imaging modalities.

METHODS: Based on the PRISMA guidelines, a comprehensive literature search of PubMed, Cochrane and Scopus databases was performed, using combinations of keywords and MeSH descriptors ('AI', 'ML', 'deep learning', 'orthopaedic surgery' and 'imaging') from inception to March 2024. Included were studies published between September 2018 and February 2024, which evaluated machine learning (ML) model effectiveness in improving orthopaedic imaging. Studies with insufficient data regarding the output variable used to assess the reliability of the ML model, those applying deterministic algorithms, unrelated topics, protocol studies, and other systematic reviews were excluded from the final synthesis. The Joanna Briggs Institute (JBI) Critical Appraisal tool and the Risk Of Bias In Non-randomised Studies-of Interventions (ROBINS-I) tool were applied for the assessment of bias among the included studies.

RESULTS: The 53 included studies reported the use of 11.990.643 images from several diagnostic instruments. A total of 39 studies reported details in terms of the Dice Similarity Coefficient (DSC), while both accuracy and sensitivity were documented across 15 studies. Precision was reported by 14, specificity by nine, and the F1 score by four of the included studies. Three studies applied the area under the curve (AUC) method to evaluate ML model performance. Among the studies included in the final synthesis, Convolutional Neural Networks (CNN) emerged as the most frequently applied category of ML models, present in 17 studies (32%).

CONCLUSION: The systematic review highlights the diverse application of AI in orthopaedic imaging, demonstrating the capability of various machine learning models in accurately segmenting and analysing orthopaedic images. The results indicate that AI models achieve high performance metrics across different imaging modalities. However, the current body of literature lacks comprehensive statistical analysis and randomized controlled trials, underscoring the need for further research to validate these findings in clinical settings.

LEVEL OF EVIDENCE: Systematic Review; Level of evidence IV.

PMID:40337671 | PMC:PMC12056712 | DOI:10.1002/jeo2.70259

Cureus. 2025 Apr 7;17(4):e81836. doi: 10.7759/cureus.81836. eCollection 2025 Apr.

ABSTRACT

This bibliometric study introduces a novel approach to assessing the application of artificial intelligence (AI) in dentistry. It analyzes trends in AI utilization across dental disciplines, treatment stages, data modalities, subsets, models, and tasks and proposes a comprehensive classification framework for AI applications in dentistry. A systematic search in the Web of Science Core Collection on December 1, 2024, using AI- and dentistry-related keywords identified original and review articles employing true AI. Data on publication details, study types, dental disciplines, treatment stages, AI subsets, models, data modalities, and tasks were extracted and analyzed using VOSviewer (Leiden University, Leiden, Netherlands) and Microsoft Excel (Microsoft Corp., Redmond, WA). Trend analysis and forecasting methods were applied to identify future research directions. Of 2,810 records, 1,368 studies met the inclusion criteria, revealing a continuous rise in AI-related dental research. While most studies focused on diagnostic applications and the orthodontics discipline, the highest recent growth was seen in treatment planning and research and education applications. Hybrid AI models and natural language processing (NLP) experienced significant increases in adoption. The most common AI tasks were classification, detection, and segmentation, although notable growth occurred in generation, data integration, and decision support. The classification framework for AI in dentistry is presented. Text-based data have shown the greatest growth among data modalities, alongside an increased use of sensor and signal data. Future research should prioritize developing NLP and hybrid AI models, conducting original studies in research and education and treatment planning, and undertaking systematic reviews focused on the diagnosis stage of prosthodontics and endodontics.

PMID:40337568 | PMC:PMC12057650 | DOI:10.7759/cureus.81836

MethodsX. 2025 Apr 24;14:103331. doi: 10.1016/j.mex.2025.103331. eCollection 2025 Jun.

ABSTRACT

The inconsistent charge and discharge patterns of electric vehicle batteries, coupled with their operation across varying voltage and current levels, pose a challenge for accurate capacity and state of health (SOH) assessment. Traditional methods rely on regular calibration, requiring controlled charge and discharge cycles, which are impractical in real-world scenarios. This research demonstrates an analysis-based method to obtain labeled capacity and SOH values in such conditions. This method not only provides labeled SOH values but also extracts health features that can be used for data-driven prediction of capacity or SOH.•Incremental capacity analysis (ICA) method has been presented to be used with electric vehicle (EV) battery data.•The approach to extract health features from a EV battery using ICA method as a function of age of the battery has been presented which can be used along with a machine learning or deep learning model.•State of health has been calculated for a vehicle battery using the proposed method.

PMID:40337556 | PMC:PMC12056397 | DOI:10.1016/j.mex.2025.103331

JHEP Rep. 2025 Mar 12;7(5):101392. doi: 10.1016/j.jhepr.2025.101392. eCollection 2025 May.

ABSTRACT

BACKGROUND & AIMS: By reducing some magnetic resonance imaging (MRI) sequences, abbreviated MRI (aMRI) has shown extensive promise for detecting hepatocellular carcinoma (HCC). We aim to develop deep learning (DL)-based gadolinium-free contrast-enhanced (CE) aMRI protocols (DL-aMRI) for detecting HCC.

METHODS: In total, 1,769 patients (913 with HCC) were retrospectively included from three institutions for training, testing, and external validation. Stable diffusion-based DL models were trained to generate CE-MRI, including T1-weighted arterial, portal venous, transitional, and hepatobiliary phase images (AP-syn, VP-syn, TP-syn, and HBP-syn, respectively). Non-contrast-MRI (NC-MRI), including T2-weighted, diffusion-weighted, and pre-contrast T1-weighted (Pre) sequences, along with either actual or DL-synthesized CE-MRI (AP, VP, TP, and HBP or AP-syn, VP-syn, TP-syn, and HBP-syn), were used to create conventional complete MRI (cMRI) and DL-aMRI protocols. An inter-method comparison of image quality between DL-aMRI and cMRI was conducted using a non-inferiority test. The sensitivity and specificity of DL-aMRI and cMRI for detecting HCC were statistically compared using the non-inferiority test and generalized estimating equations models.

RESULTS: DL-aMRI showed a remarkable reduction in acquisition time compared with cMRI (4.1 vs. 28.1 min). The image quality of DL-synthesized CE-MRI was not inferior to that of actual CE-MRI (p <0.001). There was an excellent inter-method agreement between the HCC sizes measured by the two protocols (R2 = 0.9436-0.9683). The pooled sensitivity and specificity of cMRI and DL-aMRI were 0.899 and 0.925 and 0.866 and 0.922, respectively. No significant differences were found between the sensitivity and specificity of the two protocols.

CONCLUSIONS: The proposed DL-aMRI could facilitate precise HCC diagnosis with no need for contrast agents, a substantial reduction in acquisition time, and preservation of both NC-MRI and CE-MRI data. DL-aMRI may serve as a valuable tool for HCC diagnosing.

IMPACT AND IMPLICATIONS: In this multi-center study involving 1,769 participants, we developed a generative deep learning-based abbreviated MRI (DL-aMRI) strategy that provides an efficient, contrast-agent-free alternative for detecting HCC with accuracy comparable to that of conventional complete MRI, significantly reducing acquisition time from 28.1 min to just 4.1 min. This strategy is valuable for clinicians who face significant workloads resulting from long MRI scanning times and the potential adverse effects of contrast agents, as well as for researchers focused on developing cost-effective and accessible diagnostic tools for HCC detection. The proposed DL-aMRI protocol has practical implications for clinical settings, enhancing diagnostic efficiency while maintaining high image quality, eliminating the need for contrast agents and ultimately benefiting patients and healthcare providers.

PMID:40337547 | PMC:PMC12056404 | DOI:10.1016/j.jhepr.2025.101392

Front Bioinform. 2025 Apr 23;5:1574359. doi: 10.3389/fbinf.2025.1574359. eCollection 2025.

ABSTRACT

Artificial intelligence (AI) has revolutionized numerous fields, including genomics, where it has significantly impacted variant calling, a crucial process in genomic analysis. Variant calling involves the detection of genetic variants such as single nucleotide polymorphisms (SNPs), insertions/deletions (InDels), and structural variants from high-throughput sequencing data. Traditionally, statistical approaches have dominated this task, but the advent of AI led to the development of sophisticated tools that promise higher accuracy, efficiency, and scalability. This review explores the state-of-the-art AI-based variant calling tools, including DeepVariant, DNAscope, DeepTrio, Clair, Clairvoyante, Medaka, and HELLO. We discuss their underlying methodologies, strengths, limitations, and performance metrics across different sequencing technologies, alongside their computational requirements, focusing primarily on SNP and InDel detection. By comparing these AI-driven techniques with conventional methods, we highlight the transformative advancements AI has introduced and its potential to further enhance genomic research.

PMID:40337525 | PMC:PMC12055765 | DOI:10.3389/fbinf.2025.1574359

Front Pharmacol. 2025 Apr 23;16:1594186. doi: 10.3389/fphar.2025.1594186. eCollection 2025.

ABSTRACT

Metabolic diseases, such as diabetes, pose significant risks to human health due to their complex pathogenic mechanisms, complicating the use of combination drug therapies. Natural medicines, which contain multiple bioactive components and exhibit fewer side effects, offer promising therapeutic potential. Metabolite imbalances are often closely associated with the pathogenesis of metabolic diseases. Therefore, metabolite detection not only aids in disease diagnosis but also provides insights into how natural medicines regulate metabolism, thereby supporting the development of preventive and therapeutic strategies. Deep learning has shown remarkable efficacy and precision across multiple domains, particularly in drug discovery applications. Building on this, We developed an innovative framework combining graph autoencoders (GAEs) with non-negative matrix factorization (NMF) to investigate metabolic disease pathogenesis via metabolite-disease association analysis. First, we applied NMF to extract discriminative features from established metabolite-disease associations. These features were subsequently integrated with known relationships and processed through a GAE to identify potential disease mechanisms. Comprehensive evaluations demonstrate our method's superior performance, while case studies validate its capability to reveal pathological mechanisms in metabolic disorders including diabetes. This approach may facilitate the development of natural medicine-based interventions. Our data and code are available at: https://github.com/Lqingquan/natural-medicine-discovery.

PMID:40337507 | PMC:PMC12055761 | DOI:10.3389/fphar.2025.1594186

Brain Commun. 2025 Apr 18;7(3):fcaf152. doi: 10.1093/braincomms/fcaf152. eCollection 2025.

ABSTRACT

'Brain age' is a numerical estimate of the biological age of the brain and an overall effort to measure neurodegeneration, regardless of disease type. In multiple sclerosis, accelerated brain ageing has been linked to disability accrual. Artificial intelligence has emerged as a promising tool for the assessment and quantification of the impact of neurodegenerative diseases. Despite the existence of numerous AI models, there is a noticeable lack of comparative imaging data for traditional machine learning versus deep learning in conditions such as multiple sclerosis. A retrospective observational study was initiated to analyse clinical and MRI data (4584 MRIs) from various scanners in a large longitudinal cohort (n = 1516) of people with multiple sclerosis collected from two institutions (Karolinska Institute and Oslo University Hospital) using a uniform data post-processing pipeline. We conducted a comparative assessment of brain age using a deep learning simple fully convolutional network and a well-established traditional machine learning model. This study was primarily aimed to validate the deep learning brain age model in multiple sclerosis. The correlation between estimated brain age and chronological age was stronger for the deep learning estimates (r = 0.90, P < 0.001) than the traditional machine learning estimates (r = 0.75, P < 0.001). An increase in brain age was significantly associated with higher expanded disability status scale scores (traditional machine learning: t = 5.3, P < 0.001; deep learning: t = 3.7, P < 0.001) and longer disease duration (traditional machine learning: t = 6.5, P < 0.001; deep learning: t = 5.8, P < 0.001). No significant inter-model difference in clinical correlation or effect measure was found, but significant differences for traditional machine learning-derived brain age estimates were found between several scanners. Our study suggests that the deep learning-derived brain age is significantly associated with clinical disability, performed equally well to the traditional machine learning-derived brain age measures, and may counteract scanner variability.

PMID:40337466 | PMC:PMC12056726 | DOI:10.1093/braincomms/fcaf152

Proc IEEE Int Symp Biomed Imaging. 2023 Apr;2023. doi: 10.1109/isbi53787.2023.10230423. Epub 2023 Sep 1.

ABSTRACT

Robust motion correction of fetal brain MRI slices is crucial for 3D brain volume reconstruction. However, conventional methods can only handle a limited range of motion. Hence, a deep learning model based on geometric constraints is proposed in order to predict the arbitrary motion of fetal brain MRI slices in a standard anatomical space, which consists of a global motion estimation network and a relative motion estimation network. In particular, the relative motion estimation network is used to estimate the relative motion between two adjacent slices, which is exploited as a geometric constraint. Then, sharing features between two networks make the model to learn more unique feature representations for global motion correction, and a weight-learnable strategy is employed to balance the contributions of two networks. With this design, the proposed method can estimate more complicated and large motions. Moreover, to build a large simulated fetal brain stack dataset with realistic appearance for successfully training a robust motion correction model, we introduced a control point-based method to simulate fetal motion trajectories at different gestational ages, between stacks and within 2D slices. The experimental results on a large number of fetal brain stacks demonstrate the state-of-the-art performance of our method.

PMID:40337452 | PMC:PMC12058216 | DOI:10.1109/isbi53787.2023.10230423

Front Med (Lausanne). 2025 Apr 23;12:1572750. doi: 10.3389/fmed.2025.1572750. eCollection 2025.

ABSTRACT

Myopia is a significant global health challenge, with the incidence of pathologic myopia (PM) on the rise. PM-related fundus diseases have become a leading cause of irreversible blindness. Early detection and treatment are crucial for the prevention and control of myopia. Recent advancements in artificial intelligence (AI), particularly in machine learning and deep learning algorithms, have shown promising results in the field of PM in ophthalmology. This review explores the latest developments in AI technology for managing PM, emphasizing its role in screening and diagnosis, grading and classification, and predictive assessment. AI has shown significant potential for clinical application in PM management, enhancing its intelligent, precise, and efficient practices.

PMID:40337273 | PMC:PMC12055784 | DOI:10.3389/fmed.2025.1572750

Precis Radiat Oncol. 2023 Aug 7;7(3):189-196. doi: 10.1002/pro6.1206. eCollection 2023 Sep.

ABSTRACT

PURPOSE: For postoperative vaginal brachytherapy (POVBT), the diversity of applicators complicates the creation of a generalized auto-segmentation model, and creating models for each applicator seems difficult due to the large amount of data required. We construct an auto-segmentation model of POVBT using small data via domain-adversarial neural networks (DANNs).

METHODS: CT images were obtained postoperatively from 90 patients with gynaecological cancer who underwent vaginal brachytherapy, including 60 and 30 treated with applicators A and X, respectively. A basal model was devised using data from the patients treated with applicator A; next, a DANN model was constructed using these same 60 patients as well as 10 of those treated with applicator X through transfer learning techniques. The remaining 20 patients treated with applicator X comprised the validation set. The model's performance was assessed using objective metrics and manual clinical evaluation.

RESULTS: The DANN model outperformed the basal model on both objective metrics and subjective evaluation (p<0.05 for all). The median DSC and 95HD values were 0.97 and 3.68 mm in the DANN model versus 0.94 and 5.61 mm in the basal model, respectively. Multi-centre subjective evaluation by three clinicians showed that 99%, 98%, and 81% of CT slices contoured by the DANN model were acceptable versus only 73%, 77%, and 57% of those contoured by the basal model. One clinician deemed the DANN model comparable to manual delineation.

CONCLUSION: DANNs provides a realistic approach for the wide application of automatic segmentation of POVBT and can potentially be used to construct auto-segmentation models from small datasets.

PMID:40337201 | PMC:PMC11934975 | DOI:10.1002/pro6.1206

ERJ Open Res. 2025 May 6;11(3):00784-2024. doi: 10.1183/23120541.00784-2024. eCollection 2025 May.

ABSTRACT

BACKGROUND: Interstitial lung disease (ILD) represents a heterogenous group of diseases that have substantial morbidity and mortality. The Envisia Genomic Classifier (EGC) is a test that analyses RNA derived from transbronchial biopsy (TBBx) samples to make a positive or negative genomic usual interstitial pneumonitis (UIP) designation. Our study assesses the ability for the EGC to predict progression of disease, with a longer duration of follow-up than previous studies.

METHODS: Patients referred for cryobiopsy for outpatient workup of ILD concurrently had TBBx and EGC testing performed. We performed a retrospective analysis to assess differences in progression of disease between EGC-positive and negative patients, applying Kaplan-Meier survival analysis and log-rank tests. Confidence in ILD diagnosis before and after the EGC result was also noted, and the difference in confidence levels was assessed by a Wilcoxon signed-rank test.

RESULTS: 82 patient cases were analysed. EGC-positive patients had a shorter progression-free survival (PFS) than EGC-negative patients, (p<0.0001), with 622 versus 1487 median PFS days respectively. EGC-positive patients also had worse progression in the subsets of patients with "indeterminate for UIP" computed tomography (CT) (p=0.0052), "alternative diagnosis" CT (p=0.0144) and non-idiopathic pulmonary fibrosis ILD diagnosis (p=0.0157). Additionally, EGC increased the diagnostic confidence level (p<0.0001).

CONCLUSION: EGC positivity predicts worse ILD progression over a sustained follow-up period. The ability to predict worse prediction early in the ILD course without the need for surgical biopsy would have significant clinical impact.

PMID:40337340 | PMC:PMC12053734 | DOI:10.1183/23120541.00784-2024

Development. 2025 May 1;152(9):dev204890. doi: 10.1242/dev.204890. Epub 2025 May 8.

NO ABSTRACT

PMID:40337795 | DOI:10.1242/dev.204890

Front Microbiol. 2025 Apr 23;16:1551123. doi: 10.3389/fmicb.2025.1551123. eCollection 2025.

ABSTRACT

Plant viruses participate as biotrophic parasites in complex interactions with their hosts, resulting in the regulation of a diverse range of chloroplast/mitochondria-related genes that are essential for mediating immune responses. In this study, integrative systems biology approaches were applied to identify chloroplast/mitochondrial genes during viral infections caused by a wide number of viruses in Arabidopsis thaliana, tobacco (Nicotiana tabacum L.), and rice (Oryza sativa L.). These findings indicated that 1.5% of the DEGs were common between Arabidopsis/tobacco and Arabidopsis/rice, whereas 0.1% of the DEGs were shared among all species. Approximately 90% of common DEGs are uniquely associated with chloroplasts and mitochondria in the host defense against viral infection and replication. The functions of WRKY, NAC, and MYB transcription factors in imparting resistance to viral infections can be established. Promoter analysis revealed that AP2/EREBP, DOF, and C2H2 zinc finger factors included the most frequent binding sites and played a more important role in plant-viral interactions. Comparative analysis revealed several miRNAs with defensive functions including miRNA156, miRNA160, and miRNA169. The PPI network revealed several key hub genes mostly related to chloroplasts/mitochondria, including ZAT6, CML37, CHLI, DREB, F27B13.20, and ASP2 with upregulation, also PLGG1, PSBY, APO2, POR, ERF, and CSP with downregulation. Moreover, novel hub genes with unknown functions, such as AT2G41640 and AT3G57380 have been identified. This study represents the first preliminary systems biology approach to elucidate the roles of chloroplast/mitochondria-related genes in Arabidopsis, tobacco, and rice against viral challenges by introducing valuable candidate genes for enhanced genetic engineering programs to develop virus-resistant crop varieties.

PMID:40336839 | PMC:PMC12055828 | DOI:10.3389/fmicb.2025.1551123

Brief Bioinform. 2025 May 1;26(3):bbaf201. doi: 10.1093/bib/bbaf201.

ABSTRACT

Despite the significant progress in accuracy and reliability in mass spectrometry technology, as well as the development of strategies based on isotopic labeling or internal standards in recent decades, systematic biases originating from non-biological factors remain a significant challenge in data analysis. In addition, the wide range of available normalization methods renders the choice of a suitable normalization method challenging. We systematically evaluated 17 normalization and 2 batch effect correction methods, originally developed for preprocessing DNA microarray data but widely applied in proteomics, on 6 publicly available spike-in and 3 label-free and tandem mass tag datasets. Opposed to state-of-the-art normalization practice, we found that a reduction in intragroup variation is not directly related to the effectiveness of the normalization methods. Furthermore, our results demonstrated that the methods RobNorm and Normics, specifically developed for proteomics data, in line with LoessF performed consistently well across the spike-in datasets, while EigenMS exhibited a high false-positive rate. Finally, based on experimental data, we show that normalization substantially impacts downstream analyses, and the impact is highly dataset-specific, emphasizing the importance of use-case-specific evaluations for novel proteomics datasets. For this, we developed the PROteomics Normalization Evaluator (PRONE), a unifying R package enabling comparative evaluation of normalization methods, including their impact on downstream analyses, while offering considerable flexibility, acknowledging the lack of universally accepted standards. PRONE is available on Bioconductor with a web application accessible at https://exbio.wzw.tum.de/prone/.

PMID:40336172 | DOI:10.1093/bib/bbaf201