Brief Bioinform. 2025 Jul 2;26(4):bbaf393. doi: 10.1093/bib/bbaf393.

ABSTRACT

Liquid-liquid phase separation (LLPS) is a common and important phenomenon where biomolecules form dynamic, membrane-less condensates through multivalent interactions, spontaneously separating into distinct concentration-dense and dilute phases. Research has shown that LLPS is associated with a wide range of cellular functional regulation. In this work, we establish a feature fusion framework based on a Siamese network for the prediction of LLPS proteins, which can integrate automatically extracted features from the protein itself and the protein-protein interaction (PPI) networks, and achieve good accuracy even in small sample sets. We used two representative graph embedding methods, Node2vec and DeepNF, to extract the embedding features of PPI networks and compared the impact of the two methods on model performance at different feature lengths. Our work provides a way for integrating multivalent interactions between proteins that drive LLPS, as well as a flexible framework for the fusion of different types of protein features, not only for LLPS prediction but also for other downstream prediction tasks. All relevant materials can be found at https://github.com/ispotato/SiameseNetwork_LLPS.

PMID:40794947 | DOI:10.1093/bib/bbaf393

Brief Bioinform. 2025 Jul 2;26(4):bbaf400. doi: 10.1093/bib/bbaf400.

ABSTRACT

Single-cell multi-omics technologies enable comprehensive molecular profiling, offering insights into cellular heterogeneity and biological mechanisms. However, current cross-modality translation methods struggle with high-dimensional, noisy, and sparse single-cell data. We propose single-cell Diffusion models for Cross-modality Translation (scDCT), a probabilistic framework for bidirectional cross-modality translation in single-cell data, including single-cell RNA sequencing, single-cell assay for transposase-accessible chromatin sequencing, and protein expression. scDCT integrates modality-specific autoencoders with conditional denoising diffusion probabilistic models to map inputs to latent spaces and perform probabilistic translation across modalities. This design captures cell-type heterogeneity, accounts for data sparsity, and models uncertainty during translation. Extensive experiments on eight benchmark datasets demonstrate that scDCT outperforms state-of-the-art methods across paired, unpaired, cross-type, and cross-tissue settings, offering a robust and interpretable solution for single-cell multi-omics integration.

PMID:40794946 | DOI:10.1093/bib/bbaf400

Plant Physiol. 2025 Jul 24:kiaf319. doi: 10.1093/plphys/kiaf319. Online ahead of print.

ABSTRACT

Cell expansion relies on turgor pressure and acidification-dependent loosening of the rigid cell wall. Distinct cell surface-based and intracellular auxin signaling pathways synergistically activate plasma membrane H+-ATPases, acidifying the apoplast, a prerequisite for cell elongation. Unlike in shoots, auxin inhibits cell elongation in roots. This auxin paradox highlights a largely unknown antagonistic pathway, driving root apoplast alkalinization. Auxin fluxes, regulated by the TINY ROOT HAIR 1 (TRH1)/POTASSIUM (K+) UPTAKE 4 (KUP4) transporter, modulate root gravitropism and root hair morphogenesis through the acropetal and basipetal auxin transport pathways, respectively. Here, we show that under acidic conditions, wild-type Arabidopsis (Arabidopsis thaliana) seedlings develop shorter root hairs and exhibit an agravitropic response, a defect that is even more pronounced in trh1/kup4 roots. Acidic conditions also distort auxin responses in wild-type roots, with these effects further exacerbated in trh1/kup4 roots. Remarkably, exogenous auxin application restores the trh1-like developmental defects in wild-type roots, suggesting that acidity chemiosmotically impairs active auxin transport. Advanced compartmental pH imaging combined with pharmacological applications revealed cytoplasmic and vacuolar acidification in trh1/kup4 root cells, which activates AHA2, the predominant plasma membrane H+-ATPase in roots. Proton efflux leads to apoplast acidification and rhizotoxicity, thereby inhibiting primary root elongation of trh1/kup4 seedlings. Our results demonstrate that as a proton-coupled potassium transporter, TRH1/KUP4 maintains a balance between cytosolic and apoplastic proton gradients, facilitating cytoplasm neutralization and apoplast alkalization in roots. Through this regulatory mechanism, we postulate that TRH1/KUP4 enables pH-driven intracellular auxin transport and modulates cell surface pH, driving root cell elongation and shaping root system architecture.

PMID:40795096 | DOI:10.1093/plphys/kiaf319

FEMS Microbiol Lett. 2025 Aug 6:fnaf078. doi: 10.1093/femsle/fnaf078. Online ahead of print.

ABSTRACT

Drain flies (Clogmia albipunctata) are insects that thrive in humid urban environments such as bathrooms drains and sewage systems. While their role in pathogen transmission has been suggested, little is known about their microbiome or ecology in non-clinical contexts. Using 16S rRNA gene metabarcoding, we characterized the bacterial communities of drain flies from three locations in South Korea, public bathrooms from a college in Seoul, a rural port in Ulleungdo island, and a highly frequented public park in Yeouido. In total, we obtained 221 families and 1 474 features. We found significant differences in microbiome composition and diversity as well as a small core microbiome shared among locations, with environmental bacteria such as Pseudomonas and Ralstonia being the dominant taxa across samples. The majority of the detected amplicon sequence variants (ASV) were not shared among locations. These findings suggest drain fly transport a location-specific environmental bacteria. Notably, we also identified ASVs of potential clinical relevance, including Mycobacterium, Acinetobacter baumanii, Providencia, and Nocardia. This is the first metagenomic insight into the microbiome of this species and adds to a renewed interest in the role that non-hematophagous insects play in urban microbial ecology and the spread of microbes.

PMID:40795028 | DOI:10.1093/femsle/fnaf078

Brief Bioinform. 2025 Jul 2;26(4):bbaf397. doi: 10.1093/bib/bbaf397.

ABSTRACT

Differential expression analysis provides insights into fundamental biological processes and with the advent of single-cell transcriptomics, gene expression can now be studied at the level of individual cells. Many analyses treat cells as samples and assume statistical independence. As cells are pseudoreplicates, this assumption does not hold, leading to reduced robustness, reproducibility, and an inflated type 1 error rate. In this study, we investigate various methods for differential expression analysis on single-cell data, conduct extensive benchmarking, and give recommendations for method choice. The tested methods include DESeq2, MAST, DREAM, scVI, the permutation test, distinct, and the t-test. We additionally adapt hierarchical bootstrapping to differential expression analysis on single-cell data and include it in our benchmark. We found that differential expression analysis methods designed specifically for single-cell data do not offer performance advantages over conventional pseudobulk methods such as DESeq2 when applied to individual datasets. In addition, they mostly require significantly longer run times. For atlas-level analysis, permutation-based methods excel in performance but show poor runtime, suggesting to use DREAM as a compromise between quality and runtime. Overall, our study offers the community a valuable benchmark of methods across diverse scenarios and offers guidelines on method selection.

PMID:40794957 | DOI:10.1093/bib/bbaf397

Sci Signal. 2025 Aug 12;18(899):eadr9131. doi: 10.1126/scisignal.adr9131. Epub 2025 Aug 12.

ABSTRACT

Increases in retroelement-derived double-stranded RNAs (dsRNAs) in various types of cancer cells facilitate the activation of antitumor immune responses. The long noncoding RNA EPIC1 interacts with the histone methyltransferase EZH2 and contributes to tumor immune evasion. Here, we found that EPIC1 in tumor cells suppressed cytoplasmic dsRNA accumulation, type I interferon (IFN) responses, and antitumor immunity. In various cancer cell lines, knockdown of EPIC1 stimulated the production of dsRNA from retroelements and an antiviral-like type I IFN response that activated immune cells. EPIC1 inhibited the expression of LINE, SINE, and LTR retroelements that were also repressed by EZH2, suggesting a potential role for the EPIC1-EZH2 interaction in regulating dsRNA production. In a humanized mouse model, in vivo delivery of EPIC1-targeting oligonucleotides enhanced dsRNA accumulation in breast cancer xenografts, reduced tumor growth, and increased the infiltration of T cells and inflammatory macrophages into tumors. Furthermore, EPIC1 knockdown improved the therapeutic efficacy of the immunotherapy drug pembrolizumab, a PD-1 inhibitor, in the humanized mouse model. Together, our findings establish EPIC1 as a key regulator of dsRNA-mediated type I IFN responses and highlight its potential as a therapeutic target to improve the efficacy of immunotherapy.

PMID:40794843 | DOI:10.1126/scisignal.adr9131

ISME J. 2025 Aug 6:wraf152. doi: 10.1093/ismejo/wraf152. Online ahead of print.

ABSTRACT

Soybean serves as a crucial source of plant-based protein for human diets. Recently, there is a growing incentive to extend the range of this crop to more northern latitudes, in order to enable profitable soybean production in Europe. To reach economic yields, soybean requires inoculation with symbiotic, diazotrophic rhizobial bacteria. However, the performance of commercial inocula is often variable under local conditions. Here, we present the citizen science project "Soy in 1,000 Gardens", a large-scale trapping experiment for isolating local soybean-nodulating rhizobia in Flanders, Belgium. We identified two locally isolated Bradyrhizobium strains performing at least as well as commercial strain B. diazoefficiens G49 in local field trials. Additionally, we found that nutrient content, microbial alpha diversity, and the presence of arbuscular mycorrhizal fungi in the soil were correlated with nodulation. Finally, we report a correlation between low bacterial alpha diversity and red nodule interior, and identified Tardiphaga as a dominant colonizer of red nodules.

PMID:40794807 | DOI:10.1093/ismejo/wraf152

FEBS J. 2025 Aug 12. doi: 10.1111/febs.70221. Online ahead of print.

ABSTRACT

A strong regenerative capacity is a hallmark of youth. From the tadpole's tail to the mammalian brain, young animals of many species can repair or regrow damaged tissues more effectively than older animals. Here, we take a broad perspective on ageing, inclusive of the transition from the developmental processes of embryogenesis through maturation to adulthood, as well as the processes that occur as an animal reaches the end of its lifespan. In some cases, the loss of regenerative capacity occurs once development is complete, and in others it occurs in the latter part of the animal's life. Regardless, the loss of regenerative capacity is caused by a failure to activate genes required for successful regeneration. This, in part, can be attributed to restructuring of the epigenome.

PMID:40794618 | DOI:10.1111/febs.70221

J Clin Invest. 2025 Aug 12:e179282. doi: 10.1172/JCI179282. Online ahead of print.

ABSTRACT

Despite the potential of targeted epigenetic therapies, most cancers do not respond to current epigenetic drugs. The Polycomb repressive complex EZH2 inhibitor tazemetostat was recently approved for the treatment of SMARCB1-deficient epithelioid sarcomas, based on the functional antagonism between PRC2 and SMARCB1. Through the analysis of tazemetostat-treated patient tumors, we recently defined key principles of their response and resistance to EZH2 epigenetic therapy. Here, using transcriptomic inference from SMARCB1-deficient tumor cells, we nominated the DNA damage repair kinase ATR as a target for rational EZH2 combination epigenetic therapy. We showed that EZH2 inhibition promotes DNA damage in epithelioid and rhabdoid tumor cells, at least in part via its induction of piggyBac transposable element derived 5 (PGBD5). We leveraged this collateral synthetic lethal dependency to target PGBD5-dependent DNA damage by inhibition of ATR, but not CHK1, using the ATR inhibitor elimusertib. Consequently, combined EZH2 and ATR inhibition improved therapeutic responses in diverse patient-derived epithelioid and rhabdoid tumors in vivo. This advances a combination epigenetic therapy based on EZH2-PGBD5 synthetic lethal dependency suitable for immediate translation to clinical trials for patients.

PMID:40794452 | DOI:10.1172/JCI179282

JCI Insight. 2025 Aug 12:e195981. doi: 10.1172/jci.insight.195981. Online ahead of print.

ABSTRACT

FOXP3+ Treg cells are critical for immune tolerance. Genetic deletion of the Forkhead domain containing proteins of the FOXP-subfamily member FOXP1 from Tregs results in impaired function associated with reduced CD25 expression and IL-2 signaling, but to date the only other FOXP family member expressed in Tregs, FOXP4, has been minimally studied. To investigate the potential functional interactions among FOXP family members in Treg cells, we specifically deleted Foxp1, Foxp4 or both in FOXP3+ committed Treg cells in mice. Our findings show that mice with combined, but not individual, deficiency in FOXP1 and FOXP4 exhibit lymphoproliferation, inflammation, autoimmunity, and early lethality. The combined absence of FOXP1 and FOXP4 in Tregs results in an activated/effector-like phenotype with compromised suppressive function in peripheral lymphoid organs, an enhanced germinal center response and proinflammatory cytokine production. We further show that FOXP1 and FOXP4 bind to Il2ra promoter regions to regulate CD25 expression in Tregs. Through pairwise comparison among mouse strains with Treg specific deletion of Foxp1, Foxp4 or both, our findings indicate a non-redundant but insufficient role of FOXP4 in Treg cell function.

PMID:40794436 | DOI:10.1172/jci.insight.195981

J Comput Aided Mol Des. 2025 Aug 12;39(1):65. doi: 10.1007/s10822-025-00645-w.

ABSTRACT

Eg5 is a mitotic kinesin motor protein essential for the formation of bipolar spindles during cell division. Its inhibition disrupts mitosis, leading to cell cycle arrest and apoptosis in cancer cells. This makes Eg5 a promising target for chemotherapeutic interventions, especially in cases resistant to traditional treatments. In this study, a drug repurposing strategy was employed to design and synthesise quinoline-based Schiff base derivatives as potential Eg5 inhibitors. These compounds were subjected to in vitro biological evaluations, including cytotoxicity testing against the human breast cancer cell line MDA-MB-231 and the normal mouse fibroblast cell line L929 using the MTT assay. Enzymatic assays targeting Eg5 were also conducted. Among the synthesised molecules, compound (5) demonstrated significant Eg5 inhibition in enzymatic assays, with an IC50 of 2.544 ± 0.810 µM in the Malachite Green assay and 4.03 ± 2.027 µM in the steady-state ATPase assay, and moderate inhibition against triple-negative breast cancer cells (MDA-MB-231). Computational studies, including molecular docking, molecular dynamics simulations, and MM/GBSA free energy calculations, were performed to analyse binding interactions. ADMET properties were predicted using the QikProp module. The findings suggest that targeting mitosis through Eg5 inhibition may offer a strategic approach in chemotherapy, potentially enhancing treatment efficacy.

PMID:40794230 | DOI:10.1007/s10822-025-00645-w

Assay Drug Dev Technol. 2025 Aug 12. doi: 10.1177/1540658X251365257. Online ahead of print.

ABSTRACT

From the steady stream of drug repurposing patent applications published under the Patent Cooperation Treaty (PCT), we have selected fifteen documents that first became available during the first quarter of 2025. As in each installment, some of these claims are truly surprising. Few researchers would have expected that SSRI antidepressants such as sertraline and indatraline could exhibit pronounced anticancer effects. Equally unexpected is the disclosure that sitagliptin, the first antidiabetic agent from the DPP-4 inhibitor class, may be used for the treatment of glioblastoma. Another striking example is the report that artemisinin derivatives, well known for their use against malaria, may induce differentiation in undifferentiated erythroid and myeloid cells in patients with myelodysplastic syndrome. In addition, the compound bucillamine-relatively obscure in Western medicine but long used for the treatment of rheumatoid arthritis in East Asia-has been proposed for potential benefit in organophosphate poisoning. These highlights exemplify the breadth of innovation currently shaping the drug repurposing landscape. The reviewed patent applications originate from a diverse range of jurisdictions, including France, Spain, Greece, Slovenia, South Korea, China, Japan, Canada, and the United States, illustrating the global nature of ongoing research efforts in this field.

PMID:40793958 | DOI:10.1177/1540658X251365257

J Eval Clin Pract. 2025 Aug;31(5):e70242. doi: 10.1111/jep.70242.

ABSTRACT

PURPOSE: Rare diseases (RDs), which are often chronic, degenerative, and life-threatening conditions, pose significant challenges due to their complexity and limited awareness among healthcare professionals. This study assessed the knowledge, awareness, and educational needs related to RDs among 5th- and 6th-year medical students at Atatürk University, Başkent University, and Istanbul University, as well as pediatric and non-pediatric specialists in Türkiye, with a focus on differences between these groups.

MATERIALS AND METHODS: A total of 258 physicians and 132 medical students participated. Data were collected through surveys examining demographics, self-assessed knowledge, awareness, and perceptions of RD-related education. Statistical analyses evaluated differences in knowledge and awareness across the groups.

RESULTS: Pediatric specialists reported significantly higher self-assessed RD knowledge than non-pediatric specialists. However, both groups showed notable gaps in awareness, particularly concerning RD prevalence and diagnostic timelines in Türkiye. Most participants expressed interst in further education but were unaware of available resources. Among students, 65.9% rated their RD knowledge as 'Poor' or 'Very Poor,' with no significant differences observed across institutions despite curriculum variations.

CONCLUSION: The findings highlight a critical lack of competence in RD-related knowledge among healthcare professionals, especially non-pediatric specialists. To address this gap, we recommend integrating integrating RD-specific into medical curricula, promoting continuous professional development through specialized training events, and enhancing the visibility of reliable RD information sources. These measures are crucial for improving early diagnosis and management of RDs, ultimately enhancing patient care and outcomes in Türkiye.

PMID:40793993 | DOI:10.1111/jep.70242

Clin Pharmacokinet. 2025 Aug 12. doi: 10.1007/s40262-025-01529-w. Online ahead of print.

ABSTRACT

BACKGROUND: Tacrolimus is a cornerstone of acute graft-versus-host disease (aGVHD) prophylaxis in allogeneic hematopoietic cell transplant (allo-HCT) recipients. However, a narrow therapeutic index and high interindividual variability in pharmacokinetics (PK) make starting dose selection a major challenge in clinical practice.

METHODS: Data from two PK studies conducted at the University of North Carolina Medical Center (UNCMC) were used to develop an oral tacrolimus population pharmacokinetic (popPK) model specific to adult allo-HCT recipients. Monte Carlo simulations were performed to compare the likelihood of achieving the UNCMC institutional target trough concentration range (ITR) (5-10 ng/mL) on the day of transplant (D0) under the current institutional dosing protocol, dosing recommendations from the Clinical Pharmacogenetics Implementation Consortium (CPIC), and model-derived dosing recommendations.

RESULTS: In total, 290 allo-HCT recipients contributed a total of 906 PK samples to the final analysis. A two-compartment popPK model adequately described the PK data. Population typical values of apparent clearance (TVCL/F) for 70 kg individuals receiving reduced intensity conditioning were 0.33 L/h/kg for CYP3A5 poor metabolizers (PMs) and 0.70 L/h/kg for intermediate and normal metabolizers (IMs and NMs). The probability of the population-level average D0 trough concentration being within the UNCMC ITR under the current UNCMC weight-based dosing protocol, CPIC-based, and model-derived dosing strategies were estimated to be 37%, 45%, and 76%, respectively. CYP3A5 IMs and NMs were predicted to require a 100% dose increase relative to CYP3A5 PMs.

CONCLUSIONS: We propose a new oral tacrolimus dosing strategy for adult allo-HCT recipients, which suggests the current weight-based dosing paradigm is insufficient. This new strategy includes CYP3A5 metabolizer phenotypes and conditioning regimen intensity, and could increase the percentage of allo-HCT recipients achieving target concentrations on D0.

CLINICAL TRIAL REGISTRATION NUMBER: Clinicaltrials.gov NCT04645667.

PMID:40794300 | DOI:10.1007/s40262-025-01529-w

Discov Oncol. 2025 Aug 12;16(1):1536. doi: 10.1007/s12672-025-03140-8.

ABSTRACT

PURPOSE: Omics technologies, such as genomics, transcriptomics, proteomics, and radiomics, play an increasingly important role in the diagnosis and treatment of non-small cell lung cancer (NSCLC). It is, therefore, essential to unveil the research landscape and future trends of relevant research. This study aims to explore the research fields based on omics technologies in NSCLC, elucidating the research status, hotspots, and trends from a bibliometric perspective.

METHODS: The Web of Science Core Collection was utilized to retrieve relevant publications in omics technologies and their applications in NSCLC. By using the bibliometric methods and tools ("bibliometrix" R package, VOSviewer, and CiteSpace), data and visualized analyses for annual publication outputs, countries, institutions, authors, journals, references, and keywords proceeded.

RESULTS: A total of 5,337 publications were involved in our analysis. These articles, written by 32,286 authors, originated in 5,863 institutions from 82 countries and were published in 797 journals. The Journal of Thoracic Oncology and Clinical Cancer Research were representative journals in omics-based research in NSCLC. "Survival," "adenocarcinoma," "mutation," "epidermal growth factor receptor," "resistance," and "chemotherapy" were the highest-frequency keywords. Liquid biopsy and deep learning were also trending topics in omics-related research, according to keyword clustering, trend topics, and citation burst analysis.

CONCLUSION: Omics technologies, including genomics, transcriptomics, and proteomics, were widely used in the diagnosis, prognosis, and treatment of NSCLC. And innovative methods, including liquid biopsy and deep learning, demonstrate a profound impact on advancing the understanding and treatment strategies for NSCLC and warrant further investigation.

PMID:40794364 | DOI:10.1007/s12672-025-03140-8

J Imaging Inform Med. 2025 Aug 12. doi: 10.1007/s10278-025-01630-3. Online ahead of print.

ABSTRACT

Accurate segmentation and classification of liver tumors are crucial for early diagnosis and effective treatment planning. However, conventional deep learning models such as tumor heterogeneity, class imbalance, and high computational demands face challenges, limiting their clinical deployment. This study introduces a lightweight hybrid framework combining an optimized Swin-UNet for segmentation with a Quantum Convolutional Neural Network (QCNN) for classification. The Swin-UNet is enhanced using a metaheuristic Search and Rescue (SAR) algorithm and a quadratic penalty-based objective function to balance compactness and accuracy. A Focal AUC loss function addresses class imbalance and improves sensitivity to minority regions. The QCNN leverages quantum-inspired mechanisms such as entanglement and superposition to achieve superior performance with reduced parameters. Evaluated on three benchmark datasets (3D-IRCADb, LiTS17, and MSD Task03), the framework achieves Dice scores of 85.8%, 88.7%, and 88.4%, respectively, alongside 96.7% classification accuracy. The model size is reduced to 64.16 MB, enabling real-time inference on edge devices (Jetson Nano). The QCNN classifier outperforms traditional CNNs in all metrics, demonstrating its effectiveness in high-dimensional medical data analysis. This work bridges the gap between diagnostic precision and computational efficiency, presenting a clinically viable AI solution for liver tumor analysis.

PMID:40794345 | DOI:10.1007/s10278-025-01630-3

Theor Appl Genet. 2025 Aug 12;138(9):211. doi: 10.1007/s00122-025-05003-w.

ABSTRACT

Genome-environment association (GEA) is an approach for identifying adaptive loci by combining genetic variation with environmental parameters, offering potential for improving crop resilience. However, its application to genebank accessions is limited by missing geographic origin data. To address this limitation, we explored the use of neural networks to predict the geographic origins of barley accessions and integrate imputed environmental data into GEA. Neural networks demonstrated high accuracy in cross-validation but occasionally produced ecologically implausible predictions as models solely considered geographical proximity. For example, some predicted origins were located within non-arable regions, such as the Mediterranean Sea. Using barley flowering time genes as benchmarks, GEA integrating imputed environmental data ( N = 11 , 032 ) displayed partially concordant yet complementary detection of genomic regions near flowering time genes compared to regular GEA ( N = 1 , 626 ), highlighting the potential of GEA with imputed data to complement regular GEA in uncovering novel adaptive loci. Also, contrary to our initial hypothesis anticipating a significant improvement in GEA performance by increasing sample size, our simulations yield unexpected insights. Our study suggests potential limitations in the sensitivity of GEA approaches to the considerable expansion in sample size achieved through predicting missing geographical data. Overall, our study provides insights into leveraging incomplete geographical origin data by integrating deep learning with GEA. Our findings indicate the need for further development of GEA approaches to optimize the use of imputed environmental data, such as incorporating regional GEA patterns instead of solely focusing on global associations between allele frequencies and environmental gradients across large-scale landscapes.

PMID:40794289 | DOI:10.1007/s00122-025-05003-w

EJNMMI Res. 2025 Aug 12;15(1):106. doi: 10.1186/s13550-025-01300-z.

NO ABSTRACT

PMID:40794258 | DOI:10.1186/s13550-025-01300-z

Eur Radiol Exp. 2025 Aug 12;9(1):75. doi: 10.1186/s41747-025-00612-y.

ABSTRACT

BACKGROUND: We aimed to quantify hepatic vessel volumes across chronic liver disease stages and healthy controls using deep learning-based magnetic resonance imaging (MRI) analysis, and assess correlations with biomarkers for liver (dys)function and fibrosis/portal hypertension.

METHODS: We assessed retrospectively healthy controls, non-advanced and advanced chronic liver disease (ACLD) patients using a 3D U-Net model for hepatic vessel segmentation on portal venous phase gadoxetic acid-enhanced 3-T MRI. Total (TVVR), hepatic (HVVR), and intrahepatic portal vein-to-volume ratios (PVVR) were compared between groups and correlated with: albumin-bilirubin (ALBI) and "model for end-stage liver disease-sodium" (MELD-Na) score) and fibrosis/portal hypertension (Fibrosis-4 (FIB-4) Score, liver stiffness measurement (LSM), hepatic venous pressure gradient (HVPG), platelet count (PLT), and spleen volume.

RESULTS: We included 197 subjects, aged 54.9 ± 13.8 years (mean ± standard deviation), 111 males (56.3%): 35 healthy controls, 44 non-ACLD, and 118 ACLD patients. TVVR and HVVR were highest in controls (3.9; 2.1), intermediate in non-ACLD (2.8; 1.7), and lowest in ACLD patients (2.3; 1.0) (p ≤ 0.001). PVVR was reduced in both non-ACLD and ACLD patients (both 1.2) compared to controls (1.7) (p ≤ 0.001), but showed no difference between CLD groups (p = 0.999). HVVR significantly correlated indirectly with FIB-4, ALBI, MELD-Na, LSM, and spleen volume (ρ ranging from -0.27 to -0.40), and directly with PLT (ρ = 0.36). TVVR and PVVR showed similar but weaker correlations.

CONCLUSION: Deep learning-based hepatic vessel volumetry demonstrated differences between healthy liver and chronic liver disease stages and shows correlations with established markers of disease severity.

RELEVANCE STATEMENT: Hepatic vessel volumetry demonstrates differences between healthy liver and chronic liver disease stages, potentially serving as a non-invasive imaging biomarker.

KEY POINTS: Deep learning-based vessel analysis can provide automated quantification of hepatic vascular changes across healthy liver and chronic liver disease stages. Automated quantification of hepatic vasculature shows significantly reduced hepatic vascular volume in advanced chronic liver disease compared to non-advanced disease and healthy liver. Decreased hepatic vascular volume, particularly in the hepatic venous system, correlates with markers of liver dysfunction, fibrosis, and portal hypertension.

PMID:40794249 | DOI:10.1186/s41747-025-00612-y

Int J Comput Assist Radiol Surg. 2025 Aug 12. doi: 10.1007/s11548-025-03491-1. Online ahead of print.

ABSTRACT

PURPOSE: Numerous medical procedures, such as pharmaceutical fluid injections and biopsies, require the use of a surgical needle. During such procedures, the localization of the needle is of prime importance, both to ensure that no vital organs will be or have been damaged and to confirm that the target location has been reached. The guidance to a target and its localization is done using different imaging devices, such as MRI machines, CT scans, and US devices. All of them suffer from artifacts, making the accurate localization, especially the tip, of the needle difficult. This implies the necessity for a new needle guidance technique.

METHODS: The movement of a needle through human tissue produces vibroacoustic signals which may be leveraged to retrieve information on the needle's location using data processing and deep learning techniques. We have constructed a specialized phantom with animal tissue submerged in gelatine to gather the data needed to prove this hypothesis.

RESULTS AND CONCLUSION: This paper summarizes our initial experiments, in which we preprocessed the data, converted it into two different spectrogram representations (Mel and continuous wavelet transform spectrograms), and used them as input for two different deep learning models: NeedleNet and ResNet-34. The goal of this work was to chart out an optimal direction for further research.

PMID:40794229 | DOI:10.1007/s11548-025-03491-1