Sci Rep. 2025 Apr 2;15(1):11352. doi: 10.1038/s41598-025-91394-1.

ABSTRACT

To address the challenges associated with dense and occluded targets in small target detection utilizing unmanned aerial vehicle (UAV), we propose an enhanced detection algorithm referred as the YOLOv8n-ACW. Building upon the YOLOv8n baseline network model, we have integrated Adown into the Backbone and developed a CCDHead to further improve the drone's capability to recognize small targets. Additionally, WIoU-V3 has been introduced as the loss function. Experiment results derived from the Visdrone2019 dataset indicate that, the YOLOv8n- ACW has achieved a 4.2% increase in mAP50(%) compared to the baseline model, while simultaneously reducing the parameter count by 36.7%, exhibiting superior capabilities in detecting small targets. Furthermore, utilizing a self-constructed dataset of G5-Pro drones for target detection experiments, the results indicate that the enhanced model has robust generalization capabilities in real-world environments. The UAV target detection experiment combines experimental simulation with real-world testing, while combining scientific exploration with educational objectives. This experiment has high fidelity, excellent functional scalability, and strong practicality, aiming to cultivate students' comprehensive practical and innovative abilities.

PMID:40175443 | DOI:10.1038/s41598-025-91394-1

Respir Res. 2025 Apr 2;26(1):123. doi: 10.1186/s12931-025-03195-9.

ABSTRACT

BACKGROUND: Cardiopulmonary exercise testing (CPET) is feasible, valid, reliable, and clinically useful in interstitial lung disease (ILD). However, maximal CPET values are often presented relative to body mass, whereas fat-free mass (FFM) may better reflect metabolically active muscle during exercise. Moreover, despite the value of maximal parameters, people with ILD do not always exercise maximally and therefore clinically relevant submaximal parameters must be identified. Therefore, this study assessed peak oxygen uptake (VO2peak) relative to FFM, identifying the validity of common scaling techniques; as well as characterising the oxygen uptake efficiency slope (OUES) and plateau (OUEP) as possible submaximal parameters.

METHODS: Participants with ILD underwent assessment of body composition and CPET via cycle ergometry during a single study visit. To determined effectiveness of scaling for body size, both body mass and FFM were scaled using ratio-standard (X/Y) and allometric (X/Yb) techniques. Pearsons's correlations determined agreement between OUES, OUEP, and parameters of lung function. Cohens kappa (κ) assessed agreement between OUES, OUEP and VO2peak.

RESULTS: A total of 24 participants (7 female; 69.8 ± 7.5 years; 17 with idiopathic pulmonary fibrosis) with ILD completed the study. Maximal exercise parameters did not require allometric scaling, and when scaled to FFM, it was shown that women have a significantly higher VO2peak than men (p = 0.044). Results also indicated that OUEP was significantly and positively correlated with DLCO (r = 0.719, p < 0.001), and held moderate agreement with VO2peak (κ = 0.50, p < 0.01).

CONCLUSION: This study identified that ratio-standard scaling is sufficient in removing residual effects of body size from VO2peak, and that VO2peak is higher in women when FFM is considered. Encouragingly, this study also identified OUEP as a possible alternative submaximal marker in people with ILD, and thus warrants further examination.

PMID:40176026 | DOI:10.1186/s12931-025-03195-9

Cardiovasc Diabetol. 2025 Apr 2;24(1):152. doi: 10.1186/s12933-025-02701-z.

ABSTRACT

BACKGROUND: Disruption of lipid metabolism contributes to increased cardiovascular risk in diabetes.

METHODS: We evaluated the associations between serum lipidomic profile and subclinical carotid atherosclerosis (SCA) in type 1 (T1D) and type 2 (T2D) diabetes, and in subjects without diabetes (controls) in a cross-sectional study. All subjects underwent a lipidomic analysis using ultra-high performance liquid chromatography-electrospray ionization tandem mass spectrometry, carotid ultrasound (mode B) to assess SCA, and clinical assessment. Multiple linear regression models were used to assess the association between features and the presence and burden of SCA in subjects with T1D, T2D, and controls separately. Additionally, multiple linear regression models with interaction terms were employed to determine features significantly associated with SCA within risk groups, including smoking habit, hypertension, dyslipidaemia, antiplatelet use and sex. Depending on the population under study, different confounding factors were considered and adjusted for, including sample origin, sex, age, hypertension, dyslipidaemia, body mass index, waist circumference, glycated haemoglobin, glucose levels, smoking habit, diabetes duration, antiplatelet use, and alanine aminotransferase levels.

RESULTS: A total of 513 subjects (151 T1D, 155 T2D, and 207 non-diabetic control) were included, in whom the percentage with SCA was 48.3%, 49.7%, and 46.9%, respectively. A total of 27 unique lipid species were associated with SCA in subjects with T2D, in former/current smokers with T2D, and in individuals with T2D without dyslipidaemia. Phosphatidylcholines and diacylglycerols were the main SCA-associated lipidic classes. Ten different species of phosphatidylcholines were up-regulated, while 4 phosphatidylcholines containing polyunsaturated fatty acids were down-regulated. One diacylglycerol was down-regulated, while the other 3 were positively associated with SCA in individuals with T2D without dyslipidaemia. We discovered several features significantly associated with SCA in individuals with T1D, but only one sterol could be partially annotated.

CONCLUSIONS: We revealed a significant disruption of lipid metabolism associated with SCA in subjects with T2D, and a larger SCA-associated disruption in former/current smokers with T2D and individuals with T2D who do not undergo lipid-lowering treatment.

PMID:40176064 | DOI:10.1186/s12933-025-02701-z

Biotechnol Lett. 2025 Apr 2;47(2):38. doi: 10.1007/s10529-025-03579-3.

ABSTRACT

OBJECTIVE: The aim of this study is to identify therapeutic cargos within mesenchymal stem cell (MSC)-derived small extracellular vesicles (sEVs) for the treatment of liver fibrosis, a condition that poses significant health risks.

RESULTS: sEVs from human umbilical cord-derived MSCs (UCMSCs) and their differentiated hepatocyte-like cells (hpUCMSCs) were found to alleviate liver fibrosis in mouse models, reduce fibrogenic gene expression in the liver, and inhibit hepatic stellate cell (HSC) activation, a central driver of liver fibrosis, in vitro. Deep sequencing identified differentially abundant microRNAs (miRNAs) (high-abundance: 57, low-abundance: 22) in both UCMSC- and hpUCMSC-derived sEVs, compared to HeLa cell-derived sEVs, which lack anti-liver fibrotic activity. Functional enrichment analysis of the high-abundance sEV miRNA targets revealed their involvement in transcriptional regulation, apoptosis, and cancer-related pathways, all of which are linked to liver fibrosis and hepatocellular carcinoma. Notably, many of the top 10 most abundant miRNAs reduced pro-fibrotic marker levels in activated HSCs in vitro.

CONCLUSION: The therapeutic potential of the high-abundance miRNAs shared by UCMSC- and hpUCMSC-derived sEVs in treating liver fibrosis is highlighted.

PMID:40175803 | DOI:10.1007/s10529-025-03579-3

Nat Med. 2025 Apr 2. doi: 10.1038/s41591-025-03610-0. Online ahead of print.

ABSTRACT

The human gut microbiome is most dynamic in early life. Although sweeping changes in taxonomic architecture are well described, it remains unknown how, and to what extent, individual strains colonize and persist and how selective pressures define their genomic architecture. In this study, we combined shotgun sequencing of 1,203 stool samples from 26 mothers and their twins (52 infants), sampled from childbirth to 8 years after birth, with culture-enhanced, deep short-read and long-read stool sequencing from a subset of 10 twins (20 infants) to define transmission, persistence and evolutionary trajectories of gut species from infancy to middle childhood. We constructed 3,995 strain-resolved metagenome-assembled genomes across 399 taxa, and we found that 27.4% persist within individuals. We identified 726 strains shared within families, with Bacteroidales, Oscillospiraceae and Lachnospiraceae, but not Bifidobacteriaceae, vertically transferred. Lastly, we identified weaning as a critical inflection point that accelerates bacterial mutation rates and separates functional profiles of genes accruing mutations.

PMID:40175737 | DOI:10.1038/s41591-025-03610-0

Nature. 2025 Apr 2. doi: 10.1038/s41586-025-08814-5. Online ahead of print.

ABSTRACT

Anthropogenic biodiversity decline threatens the functioning of ecosystems and the many benefits they provide to humanity1. As well as causing species losses in directly affected locations, human influence might also reduce biodiversity in relatively unmodified vegetation if far-reaching anthropogenic effects trigger local extinctions and hinder recolonization. Here we show that local plant diversity is globally negatively related to the level of anthropogenic activity in the surrounding region. Impoverishment of natural vegetation was evident only when we considered community completeness: the proportion of all suitable species in the region that are present at a site. To estimate community completeness, we compared the number of recorded species with the dark diversity-ecologically suitable species that are absent from a site but present in the surrounding region2. In the sampled regions with a minimal human footprint index, an average of 35% of suitable plant species were present locally, compared with less than 20% in highly affected regions. Besides having the potential to uncover overlooked threats to biodiversity, dark diversity also provides guidance for nature conservation. Species in the dark diversity remain regionally present, and their local populations might be restored through measures that improve connectivity between natural vegetation fragments and reduce threats to population persistence.

PMID:40175550 | DOI:10.1038/s41586-025-08814-5

Sci Rep. 2025 Apr 2;15(1):11312. doi: 10.1038/s41598-025-95276-4.

ABSTRACT

Infectious mononucleosis (IM), predominantly caused by primary Epstein-Barr virus (EBV) infection, is a common disease in adolescents and young adults. EBV infection is nearly ubiquitous globally. Although primary EBV infection is asymptomatic in most individuals, IM manifests in a subset infected during adolescence or young adulthood. IM occurrence is linked to sibship structure, and is associated with increased risk of multiple sclerosis, other autoimmune diseases, and cancer later in life. We analyzed 47 biomarkers in 5,526 Danish individuals aged 18-60 years, of whom 604 had a history of IM, examining their associations with IM history up to 48 years after IM diagnosis. No significant long-term associations were observed after adjusting for multiple comparisons. When restricting the analysis to individuals measured within 10 years post-IM diagnosis, a statistically significant increase in CRP levels was observed in females. This association was not driven by oral contraceptive use. No significant associations between sibship structure and biomarker levels were detected. In conclusion, our study shows that while IM may lead to a transient increase in CRP levels in females, it does not result in long-term alterations in plasma biomarkers related to immune function, suggesting other mechanisms may be responsible for the long-term health impacts associated with IM.

PMID:40175486 | DOI:10.1038/s41598-025-95276-4

Microsyst Nanoeng. 2025 Apr 2;11(1):56. doi: 10.1038/s41378-025-00880-z.

ABSTRACT

Chronic kidney disease (CKD) significantly affects people's health and quality of life and presents a high economic burden worldwide. There are well-established biomarkers for CKD diagnosis. However, the existing routine standard tests are lab-based and governed by strict regulations. Creatinine is commonly measured as a filtration biomarker in blood to determine estimated Glomerular Filtration Rate (eGFR), as well as a normalization factor to calculate urinary Albumin-to-Creatinine Ratio (uACR) for CKD evaluation. In this study, we developed a passive flow microreactor for colorimetric urine creatinine measurement (uCR-Chip), which is highly amenable to integration with our previously developed microfluidic urine albumin assay. The combination of the 2-phase pressure compensation (2-PPC) technique and microfluidic channel network design accurately controls the fluidic mixing ratio and chemical reaction. Together with an optimized observation window (OW) design, a uniform and stable detection signal was achieved within 7 min. The color signal was measured by a simple USB microscope-based platform to quantify creatinine concentration in the sample. The combination of the custom in-house photomask production techniques and dry-film photoresist-based lithography enabled rapid iterative design optimization and precise chip fabrication. The developed assay achieved a dynamic linear detection range up to 40 mM and a lower limit of detection (LOD) of 0.521 mM, meeting the clinical precision requirements (comparable to existing point-of-care (PoC) systems). The microreactor was validated using creatinine standards spiked into commercial artificial urine that mimics physiological matrix. Our results showed acceptable recovery rate and low matrix effect, especially for the low creatinine concentration range in comparison to a commercial PoC uACR test. Altogether, the developed uCR-Chip offers a viable PoC test for CKD assessment and provides a potential platform technology to measure various disease biomarkers.

PMID:40175342 | DOI:10.1038/s41378-025-00880-z

Trends Plant Sci. 2025 Apr 1:S1360-1385(25)00063-9. doi: 10.1016/j.tplants.2025.03.003. Online ahead of print.

ABSTRACT

The field of mass spectrometry (MS)-based proteomics is rapidly advancing with technological and computational improvements, including leveraging the power of artificial intelligence (AI) to drive innovation. Such innovation has been particularly apparent in human disease research, where the intersection of these disciplines has pioneered a new age of disease diagnostics and pharmaceutical discovery. However, applications within plant sciences remains woefully under-represented and yet provides exceptional promise and potential to support new, interdisciplinary areas of research. Timely and novel examples of proteomics advancing plant science encompass biotechnology, climatic resiliency, agricultural production systems, and disease management. Herein, we propose new scientific avenues that leverage the power of proteomics and AI within plant science research to drive new discoveries and innovation.

PMID:40175191 | DOI:10.1016/j.tplants.2025.03.003

Adv Pharmacol. 2025;103:287-311. doi: 10.1016/bs.apha.2025.01.012. Epub 2025 Feb 26.

ABSTRACT

Bioinformatics has taken a pivotal place in the life sciences field. Not only does it improve, but it also fine-tunes and complements the wet lab experiments. It has been a driving force in the so-called biological sciences, converting them into hypothesis and data-driven fields. This study highlights the translational impact of bioinformatics on experimental biology and discusses its evolution and the advantages it has brought to advancing biological research. Computational analyses make labor-intensive wet lab work cost-effective by reducing the use of expensive reagents. Genome/proteome-wide studies have become feasible due to the efficiency and speed of bioinformatics tools, which can hardly be compared with wet lab experiments. Computational methods provide the scalability essential for manipulating large and complex data of biological origin. AI-integrated bioinformatics studies can unveil important biological patterns that traditional approaches may otherwise overlook. Bioinformatics contributes to hypothesis formation and experiment design, which is pivotal for modern-day multi-omics and systems biology studies. Integrating bioinformatics in the experimental procedures increases reproducibility and helps reduce human errors. Although today's AI-integrated bioinformatics predictions have significantly improved in accuracy over the years, wet lab validation is still unavoidable for confirming these predictions. Challenges persist in multi-omics data integration and analysis, AI model interpretability, and multiscale modeling. Addressing these shortcomings through the latest developments is essential for advancing our knowledge of disease mechanisms, therapeutic strategies, and precision medicine.

PMID:40175046 | DOI:10.1016/bs.apha.2025.01.012

Adv Pharmacol. 2025;103:231-263. doi: 10.1016/bs.apha.2025.01.003. Epub 2025 Feb 6.

ABSTRACT

Computational precision in drug discovery integrates algorithms and high-performance computing to analyze complex biological data with unprecedented accuracy, revolutionizing the identification of therapeutic targets. This process encompasses diverse computational and experimental approaches that enhance drug discovery's speed and precision. Advanced techniques like next-generation sequencing enable rapid genetic characterization, while proteomics explores protein expression and interactions driving disease progression. In-silico methods, including molecular docking, virtual screening, and pharmacophore modeling, predict interactions between small molecules and biological targets, accelerating early drug candidate identification. Structure-based drug design and molecular dynamics simulations refine drug designs by elucidating target structures and molecular behaviors. Ligand-based methods utilize known chemical properties to anticipate new compound activities. AI and machine learning optimizes data analysis, offering novel insights and improving predictive accuracy. Systems biology and network pharmacology provide a holistic view of biological networks, identifying critical nodes as potential drug targets, which traditional methods might overlook. Computational tools synergize with experimental techniques, enhancing the treatment of complex diseases with personalized medicine by tailoring therapies to individual patients. Ethical and regulatory compliance ensures clinical applicability, bridging computational predictions to effective therapies. This multi-dimensional approach marks a paradigm shift in modern medicine, delivering safer, more effective treatments with precision. By integrating bioinformatics, genomics, and proteomics, computational drug discovery has transformed how therapeutic interventions are developed, ensuring an era of personalized, efficient healthcare.

PMID:40175044 | DOI:10.1016/bs.apha.2025.01.003

Adv Pharmacol. 2025;103:1-22. doi: 10.1016/bs.apha.2025.01.006. Epub 2025 Feb 13.

ABSTRACT

In the current scenario of pandemics, drug discovery and design have undergone a significant transformation due to the integration of advanced computational methodologies. These methodologies utilize sophisticated algorithms, machine learning, artificial intelligence, and high-performance computing to expedite the drug development process, enhances accuracy, and reduces costs. Machine learning and AI have revolutionized predictive modeling, virtual screening, and de novo drug design, allowing for the identification and optimization of novel compounds with desirable properties. Molecular dynamics simulations provide a detailed insight into protein-ligand interactions and conformational changes, facilitating an understanding of drug efficacy at the atomic level. Quantum mechanics/molecular mechanics methods offer precise predictions of binding energies and reaction mechanisms, while structure-based drug design employs docking studies and fragment-based design to improve drug-receptor binding affinities. Network pharmacology and systems biology approaches analyze polypharmacology and biological networks to identify novel drug targets and understand complex interactions. Cheminformatics explores vast chemical spaces and employs data mining to find patterns in large datasets. Computational toxicology predicts adverse effects early in development, reducing reliance on animal testing. Bioinformatics integrates genomic, proteomic, and metabolomics data to discover biomarkers and understand genetic variations affecting drug response. Lastly, cloud computing and big data technologies facilitate high-throughput screening and comprehensive data analysis. Collectively, these computational innovations are driving a paradigm shift in drug discovery and design, making it more efficient, accurate, and cost-effective.

PMID:40175036 | DOI:10.1016/bs.apha.2025.01.006

J Am Coll Radiol. 2025 Mar 31:S1546-1440(25)00195-4. doi: 10.1016/j.jacr.2025.03.006. Online ahead of print.

NO ABSTRACT

PMID:40174871 | DOI:10.1016/j.jacr.2025.03.006

Nephron. 2025 Apr 2:1-11. doi: 10.1159/000544754. Online ahead of print.

ABSTRACT

BACKGROUND: Although there have been reports of relapse or worsening of membranous nephropathy after receiving vaccines against coronavirus disease 2019 (COVID-19), the causal relationship or association between them has not been established. This study aimed to investigate the occurrence of relapse or worsening of membranous nephropathy following inactivated COVID-19 vaccination.

METHODS: Patients who had been diagnosed with membranous nephropathy before receiving their first dose of vaccination, or before March 1, 2021, for unvaccinated patients, were included in the study. All patients were monitored at the Membranous Nephropathy Clinic of Huashan Hospital, Fudan University. The reasons for not receiving vaccines were investigated. The impact of COVID-19 vaccination on membranous nephropathy was assessed by comparing the relapse or worsening of membranous nephropathy within 12 months in vaccinated and unvaccinated patients with proteinuria <3.5 g/d. The baseline variables were balanced using cardinality matching.

RESULTS: A total of 353 patients with membranous nephropathy were included in the study, with 186 (53%) having received inactivated COVID-19 vaccines. Among the 167 unvaccinated participants, 114 (68%) expressed concerns about the possibility of disease relapse, and 47 (28%) were worried about the vaccine's efficacy due to their immunosuppressive therapy. Of the 239 participants with proteinuria <3.5 g/d, 152 were vaccinated, and 16 (11%) experienced a relapse or worsening of the disease during the follow-up period, which was similar to the 14 (16%) observed in the unvaccinated group. Following cardinality matching, there was no difference in the rate of relapse or worsening between the two groups, with 10 (13%) in the vaccinated group and 11 (15%) in the unvaccinated group (hazard ratio 0.98, 95% confidence interval 0.42-2.33).

CONCLUSION: Getting the inactivated COVID-19 vaccine may not increase risk of relapse or worsening in patients with membranous nephropathy.

PMID:40174580 | DOI:10.1159/000544754

Reprod Biomed Online. 2024 Nov 14;50(5):104707. doi: 10.1016/j.rbmo.2024.104707. Online ahead of print.

ABSTRACT

Infertility, adverse pregnancy outcomes and genital infections are global concerns. The reproductive tract microbiome appears to play a crucial role in the physiology of both the female and male reproductive tracts. Despite the presence of thousands of microbes in body fluids shared during unprotected sexual intercourse, they have traditionally been studied separately, with greater emphasis on the female (mostly vaginal) microbiome, and the interaction between these microbiomes in a sexually active couple has been overlooked. This review introduces the concept of the 'seminovaginal microbiome' - the collective microbiota of both partners, transferred and shared during sexual interaction. By synthesizing the existing body of next-generation sequencing-based literature, this review establishes the first holistic view of how these microbiomes interact, influence reproductive health and affect assisted reproductive technique outcomes, as well as the occurrence of microbe-associated diseases such as sexually transmitted infections, prostatitis, bacterial vaginosis and candidiasis. Additionally, the microbial interplay in homosexual couples and transgender individuals is discussed.

PMID:40174296 | DOI:10.1016/j.rbmo.2024.104707

PLoS Comput Biol. 2025 Apr 2;21(4):e1012876. doi: 10.1371/journal.pcbi.1012876. Online ahead of print.

ABSTRACT

Computational approaches for drug repurposing for viral diseases have mainly focused on a small number of antivirals that directly target pathogens (virus centric therapies). In this work, we combine ideas from collaborative filtering and network medicine for making predictions on a much larger set of drugs that could be repurposed for host centric therapies, that are aimed at interfering with host cell factors required by a pathogen. Our idea is to create matrices quantifying the perturbation that drugs and viruses induce on human protein interaction networks. Then, we decompose these matrices to learn embeddings of drugs, viruses, and proteins in a low dimensional space. Predictions of host-centric antivirals are obtained by taking the dot product between the corresponding drug and virus representations. Our approach is general and can be applied systematically to any compound with known targets and any virus whose host proteins are known. We show that our predictions have high accuracy and that the embeddings contain meaningful biological information that may provide insights into the underlying biology of viral infections. Our approach can integrate different types of information, does not rely on known drug-virus associations and can be applied to new viral diseases and drugs.

PMID:40173200 | DOI:10.1371/journal.pcbi.1012876

PLoS Pathog. 2025 Apr 2;21(4):e1013021. doi: 10.1371/journal.ppat.1013021. Online ahead of print.

ABSTRACT

Pseudomonas aeruginosa is an opportunistic pathogen responsible for airway infections in immunocompromised individuals, including those with cystic fibrosis (CF). P. aeruginosa has a type III secretion system (T3SS) that translocates effectors into host cells. ExoS is a T3SS effector with ADP ribosyltransferase (ADPRT) activity. ExoS ADPRT activity promotes P. aeruginosa virulence by inhibiting phagocytosis and limiting oxidative burst in neutrophils. The P. aeruginosa T3SS also translocates flagellin, which can activate the NLRC4 inflammasome, resulting in: 1) gasdermin-D pores, release of IL-1β and pyroptosis; and 2) histone 3 citrullination (CitH3), nuclear DNA decondensation and expansion into the neutrophil cytosol with incomplete NET extrusion. However, studies with P. aeruginosa PAO1 indicate that ExoS ADPRT activity inhibits the NLRC4 inflammasome in neutrophils. Here, we identified an ExoS+ CF clinical isolate of P. aeruginosa with a hyperactive T3SS. Variants of the hyperactive T3SS mutant or PAO1 were used to infect neutrophils from C57BL/6 mice that were wildtype or engineered to have a CF genotype or defects in inflammasome assembly. Responses to NLRC4 inflammasome assembly or ExoS ADPRT activity were assayed and found to be similar for C57BL/6 or CF neutrophils. ExoS ADPRT activity in the hyperactive T3SS mutant regulated inflammasome, nuclear DNA decondensation and incomplete NET extrusion responses, like PAO1, but promoted enhanced CitH3 and plasma membrane rupture (PMR). Glycine supplementation inhibited PMR by the hyperactive T3SS mutant, suggesting ninjurin-1 is required for this process. These results identify enhanced neutrophil PMR as a pathogenic activity of ExoS ADPRT in hypervirulent P. aeruginosa.

PMID:40173191 | DOI:10.1371/journal.ppat.1013021

Invest Radiol. 2025 Apr 2. doi: 10.1097/RLI.0000000000001182. Online ahead of print.

ABSTRACT

OBJECTIVES: Respiratory motion poses a significant challenge for clinical workflows in diagnostic imaging and radiation therapy. Many technologies such as motion artifact reduction and tumor tracking have been developed to compensate for its effect. To assess these technologies, respiratory motion phantoms (RMPs) are required as preclinical testing environments, for instance, in computed tomography (CT). However, current CT RMPs are highly simplified and do not exhibit realistic tissue structures or deformation patterns. With the rise of more complex motion compensation technologies such as deep learning-based algorithms, there is a need for more realistic RMPs. This work introduces PixelPrint4D, a 3D printing method for fabricating lifelike, patient-specific deformable lung phantoms for CT imaging.

MATERIALS AND METHODS: A 4DCT dataset of a lung cancer patient was acquired. The volumetric image data of the right lung at end inhalation was converted into 3D printer instructions using the previously developed PixelPrint software. A flexible 3D printing material was used to replicate variable densities voxel-by-voxel within the phantom. The accuracy of the phantom was assessed by acquiring CT scans of the phantom at rest, and under various levels of compression. These phantom images were then compiled into a pseudo-4DCT dataset and compared to the reference patient 4DCT images. Metrics used to assess the phantom structural accuracy included mean attenuation errors, 2-sample 2-sided Kolmogorov-Smirnov (KS) test on histograms, and structural similarity index (SSIM). The phantom deformation properties were assessed by calculating displacement errors of the tumor and throughout the full lung volume, attenuation change errors, and Jacobian errors, as well as the relationship between Jacobian and attenuation changes.

RESULTS: The phantom closely replicated patient lung structures, textures, and attenuation profiles. SSIM was measured as 0.93 between the patient and phantom lung, suggesting a high level of structural accuracy. Furthermore, it exhibited realistic nonrigid deformation patterns. The mean tumor motion errors in the phantom were ≤0.7 ± 0.6 mm in each orthogonal direction. Finally, the relationship between attenuation and local volume changes in the phantom had a strong correlation with that of the patient, with analysis of covariance yielding P = 0.83 and f = 0.04, suggesting no significant difference between the phantom and patient.

CONCLUSIONS: PixelPrint4D facilitates the creation of highly realistic RMPs, exceeding the capabilities of existing models to provide enhanced testing environments for a wide range of emerging CT technologies.

PMID:40173424 | DOI:10.1097/RLI.0000000000001182

Aesthet Surg J. 2025 Apr 2:sjaf047. doi: 10.1093/asj/sjaf047. Online ahead of print.

ABSTRACT

BACKGROUND: Breast Implant Illness (BII) is a spectrum of symptoms some people attribute to breast implants. While causality remains unproven, patient interest has grown significantly. Understanding patient perceptions of BII on social media is crucial as these platforms increasingly influence healthcare decisions.

OBJECTIVES: The purpose of this study is to analyze patient perceptions and emotional responses to BII on social media using RoBERTa, a natural processing model trained on 124 million X posts.

METHODS: Posts mentioning BII from 2014-2023 were analyzed using two NLP models: one for sentiment (positive/negative) and another for emotions (fear, sadness, anger, disgust, neutral, surprise, and joy). Posts were then classified by their highest-scoring emotion. Results were compared over across 2014-2018 and 2019-2023, with correlation analysis (Pearson correlation coefficient) between published implant explantation and augmentation data.

RESULTS: Analysis of 6,099 posts over 10 years showed 75.4% were negative, with monthly averages of 50.85 peaking at 213 in March 2019. Fear and neutral emotions dominated, representing 35.9% and 35.6% respectively. The strongest emotions were neutral and fear, with an average score of 0.293 and 0.286 per post, respectively. Fear scores increased from 0.219 (2014-2018) to 0.303 (2019-2023). Strong positive correlations (r>0.70) existed between annual explantation rates/explantation-to-augmentation ratios and total, negative, neutral, and fear posts.

CONCLUSIONS: BII discourse on X peaked in 2019, characterized predominantly by negative sentiment and fear. The strong correlation between fear/negative-based posts and explantation rates suggests social media discourse significantly influences patient decisions regarding breast implant removal.

PMID:40173420 | DOI:10.1093/asj/sjaf047

PLoS One. 2025 Apr 2;20(4):e0319078. doi: 10.1371/journal.pone.0319078. eCollection 2025.

ABSTRACT

Hepatitis is a widespread inflammatory condition of the liver, presenting a formidable global health challenge. Accurate and timely detection of hepatitis is crucial for effective patient management, yet existing methods exhibit limitations that underscore the need for innovative approaches. Early-stage detection of hepatitis is now possible with the recent adoption of machine learning and deep learning approaches. With this in mind, the study investigates the use of traditional machine learning models, specifically classifiers such as logistic regression, support vector machines (SVM), decision trees, random forest, multilayer perceptron (MLP), and other models, to predict hepatitis infections. After extensive data preprocessing including outlier detection, dataset balancing, and feature engineering, we evaluated the performance of these models. We explored three modeling approaches: machine learning with default hyperparameters, hyperparameter-tuned models using GridSearchCV, and ensemble modeling techniques. The SVM model demonstrated outstanding performance, achieving 99.25% accuracy and a perfect AUC score of 1.00 with consistency in other metrics with 99.27% precision, and 99.24% for both recall and F1-measure. The MLP and Random Forest proved to be in pace with the superior performance of SVM exhibiting an accuracy of 99.00%. To ensure robustness, we employed a 5-fold cross-validation technique. For deeper insight into model interpretability and validation, we employed an explainability analysis of our best-performed models to identify the most effective feature for hepatitis detection. Our proposed model, particularly SVM, exhibits better prediction performance regarding different performance metrics compared to existing literature.

PMID:40173410 | DOI:10.1371/journal.pone.0319078