Cell Death Dis. 2025 Mar 21;16(1):196. doi: 10.1038/s41419-025-07522-2.

ABSTRACT

The intracellular enzyme 11β-hydroxysteroid dehydrogenase type 1 (11βHSD1) catalyzes the interconversion of active glucocorticoid (cortisol) and its intrinsically inert form (cortisone) in metabolic tissues. Although 11βHSD1 is considered a promising therapeutic target in metabolic disorders such as type 2 diabetes, obesity, and nonalcoholic steatohepatitis because of its hepatic functions, its roles in other tissues have received less attention. In this study, we show that the 11βHSD1-specific inhibitor J2H-1702 facilitates the reversion of endothelial-to-mesenchymal transition in multicellular lung spheroid models encapsulating the complex crosstalk among lung cancer cells, vascular endothelial cells, and macrophages. In vascular endothelial cells, J2H-1702 not only suppressed interleukin-1α (IL-1α) expression but also attenuated reactive oxygen species-induced DNA damage by upregulating heme oxygenase-1. Additionally, in macrophages, which are key regulators of fibrogenesis, inhibition of 11βHSD1 markedly reduced IL-1β expression, thereby modulating the pro-inflammatory phenotype of activated macrophages. In mouse models of pulmonary fibrosis, including a bleomycin-induced idiopathic model and a radiation-induced model, J2H-1702 alleviated pulmonary fibrosis and markedly improved the efficacy of nintedanib. Collectively, our data suggest that J2H-1702 holds promise as a clinical candidate for the treatment of pulmonary fibrosis associated with reactive oxygen species-induced DNA damage, endothelial-to-mesenchymal transition, and inflammatory responses.

PMID:40118823 | DOI:10.1038/s41419-025-07522-2

New Phytol. 2025 Mar 21. doi: 10.1111/nph.70063. Online ahead of print.

ABSTRACT

Plant disease outbreaks, exacerbated by climate change, threaten food security and environmental sustainability world-wide. Plants interact with a wide range of microorganisms. The quest for resilient agriculture requires a deep insight into the molecular and ecological interplays between plants and their associated microbial communities. Omics methods, by profiling entire molecular sets, have shed light on these complex interactions. Nonetheless, deciphering the relationships among thousands of molecular components remains a formidable challenge, and studies that integrate these components into cohesive biological networks involving plants and associated microbes are still limited. Systems biology has the potential to predict the effects of biotic and abiotic perturbations on these networks. It is therefore a promising framework for addressing the full complexity of plant-microbiome interactions.

PMID:40119556 | DOI:10.1111/nph.70063

Rice (N Y). 2025 Mar 22;18(1):20. doi: 10.1186/s12284-025-00774-8.

ABSTRACT

Grain size is vital determinant for grain yield and quality, which specified by its three-dimensional structure of seeds (length, width and thickness). The ZINC FINGER-HOMEODOMAIN (ZHD) proteins play critical roles in plant growth and development. However, the information regarding the function in reproductive development of ZHD proteins is scarce. Here, we deeply characterized the phenotype of oszhd1, oszhd2, and oszhd1oszhd2. The single mutants of OsZHD1/2 were similar with wild type. Nevertheless, the double mutant displayed dwarfism and smaller reproductive organs, and shorter, narrower, and thinner grain size. oszhd1oszhd2 revealed a significant decrease in total cell length and number, and single cell width in outer parenchyma; reducing the average width of longitudinal epidermal cells, but the length were increased in outer and inner glumes of oszhd1oszhd2 compared with wild-type, oszhd1-1, oszhd2-1, respectively. OsZHD1 and OsZHD2 encoded the nucleus protein and were distributed predominately in stem and the developing spikelets, asserting their roles in grain size. Meanwhile, yeast two-hybrid, bimolecular fluorescence complementation, and Co-immunoprecipitation assay clarified that OsZHD1 could directly interacted with OsZHD2. The differential expression analysis showed that 839 DEGs, which were down-regulated in oszhd1oszhd2 than wild type and single mutants, were mainly enriched in secondary metabolite biosynthetic, integral component of membrane, and transporter activity pathway. Moreover, it is reliable that the altered expression of cell cycle and expansion-related and grain size-related genes were observed in RNA-seq data, highly consistent with the qRT-PCR results. Altogether, our results suggest that OsZHD1/2 are functional redundancy and involved in regulating grain size by influencing cell proliferation in rice.

PMID:40119214 | DOI:10.1186/s12284-025-00774-8

Cell Res. 2025 Mar 21. doi: 10.1038/s41422-025-01095-7. Online ahead of print.

ABSTRACT

Biased allosteric modulators (BAMs) of G protein-coupled receptors (GPCRs) have been at the forefront of drug discovery owing to their potential to selectively stimulate therapeutically relevant signaling and avoid on-target side effects. Although structures of GPCRs in complex with G protein or GRK in a BAM-bound state have recently been resolved, revealing that BAM can induce biased signaling by directly modulating interactions between GPCRs and these two transducers, no BAM-bound GPCR-arrestin complex structure has yet been determined, limiting our understanding of the full pharmacological profile of BAMs. Herein, we developed a chemical protein synthesis strategy to generate neurotensin receptor 1 (NTSR1) with defined hexa-phosphorylation at its C-terminus and resolved high-resolution cryo-EM structures (2.65-2.88 Å) of NTSR1 in complex with both β-arrestin1 and the BAM SBI-553. These structures revealed a unique "loop engagement" configuration of β-arrestin1 coupling to NTSR1 in the presence of SBI-553, markedly different from the typical "core engagement" configuration observed in the absence of BAMs. This configuration is characterized by the engagement of the intracellular loop 3 of NTSR1 with a cavity in the central crest of β-arrestin1, representing a previously unobserved, arrestin-selective conformation of GPCR. Our findings fill the critical knowledge gap regarding the regulation of GPCR-arrestin interactions and biased signaling by BAMs, which would advance the development of safer and more efficacious GPCR-targeted therapeutics.

PMID:40118988 | DOI:10.1038/s41422-025-01095-7

Sci Rep. 2025 Mar 21;15(1):9768. doi: 10.1038/s41598-025-94632-8.

ABSTRACT

Because the innate immunity might and fungi in the lungs might enhance the severity of lupus-induced diffuse alveolar hemorrhage (DAH), intraperitoneal pristane injection was performed in C57BL6 mice with intratracheal administration by Candida albicans or phosphate buffer solution (PBS). Despite the similar pristane-induced lupus (proteinuria, serum creatinine, and serum anti-dsDNA) at 5 weeks of the model, Candida administration worsened several characteristics, including mortality, body weight, serum cytokines (TNF-α and IL-6), and lung hemorrhage score, and cytokines in the lung tissue (TNF-α, IL-6, and IL-10), but not gut permeability (FITC-dextran assay), serum IL-10, immune cells in the spleens (flow cytometry analysis), and activities of peritoneal macrophages (polymerase-chain reaction). Although Candida administration reduced proteobacterial abundance and altered alpha and beta diversity compared with PBS control, lung microbiota was not different between Candida administration in pristane- and non-pristane-administered mice. Because of the prominent Gram-negative bacteria in lung microbiota and the role of neutrophils in DAH, lipopolysaccharide (LPS) with and without heat-killed Candida preparation was tested. Indeed, Candida preparation with LPS induced more severe pro-inflammatory neutrophils than LPS stimulation alone as indicated by the expression of several genes (TNF-α, IL-6, IL-1β, IL-10, Dectin-1, and NF-κB). In conclusion, the intratracheal Candida worsened pristane-induced lung hemorrhage partly through the enhanced neutrophil responses against bacteria and fungi. More studies on Candida colonization in sputum from patients with lupus-induced DAH are interesting.

PMID:40118938 | DOI:10.1038/s41598-025-94632-8

Nat Commun. 2025 Mar 21;16(1):2223. doi: 10.1038/s41467-025-57345-0.

ABSTRACT

The delay between the origin of animals in the Neoproterozoic and their Cambrian diversification remains perplexing. Animal diversification mirrors an expansion in marine shelf area under a greenhouse climate, though the extent to which these environmental conditions directly influenced physiology and early organismal ecology remains unclear. Here, we use a biogeochemical model to quantify oxygen dynamics at the sunlit sediment-water interface over day-night (diel) cycles at warm and cold conditions. We find that warm temperatures dictated physiologically stressful diel benthic oxic-anoxic shifts over a nutrient-rich shelf. Under these conditions, a population-and-phenotype model further show that the benefits of efficient cellular oxygen sensing that can offer adaptations to stress outweigh its cost. Since diurnal benthic redox variability would have expanded as continents were flooded in the end-Neoproterozoic and early Palaeozoic, we propose that a combination of physiological stress and ample resources in the benthic environment may have impacted the adaptive radiation of animals tolerant to oxygen fluctuations.

PMID:40118825 | DOI:10.1038/s41467-025-57345-0

Geriatr Gerontol Int. 2025 Mar 21. doi: 10.1111/ggi.70035. Online ahead of print.

ABSTRACT

AIM: The use of potentially inappropriate medications (PIMs) in older adults can increase the risk of drug-related adverse events. We aimed to examine the associations between PIMs, frailty, and each frailty component in community-dwelling older women.

METHODS: This cross-sectional study included participants aged ≥65 years from a prospective cohort of older Japanese women. Frailty was classified using the Japanese version of Fried's Frailty Criteria, comprising five components. PIMs were identified using a screening tool for Japanese among regular prescription medications collected from participants' prescription notebooks. Multivariable logistic regression models adjusted for age and comorbidities were used to examine the association between PIMs (0, 1, 2, ≥3), frailty, and each component. The possible interactions between age groups (65-74 and ≥75 years) and PIMs were investigated. Age-stratified analyses were also performed.

RESULTS: We analyzed 530 older women (median age [interquartile range], 71 [68, 75] years) with a frailty prevalence of 5.5%. Three or more PIMs were associated with frailty and weight loss (adjusted odds ratio [95% confidence interval], 3.80 [1.23, 11.80], 2.53 [1.15, 5.39]). In age-stratified analyses, ≥3 PIMs were associated with weight loss (8.39 [1.79, 48.98]) in women aged ≥75 years, whereas 1 or 2 PIMs were associated with frailty (4.52 [1.17, 19.08]) or weakness (3.13 [1.22, 7.78]) in those aged 65-74 years.

CONCLUSIONS: Our results may suggest that the number of PIM prescriptions is associated with frailty and frailty components in older women. Longitudinal studies are required to clarify the causality between the number of PIMs and frailty. Geriatr Gerontol Int 2025; ••: ••-••.

PMID:40119543 | DOI:10.1111/ggi.70035

Neurotherapeutics. 2025 Mar 20:e00575. doi: 10.1016/j.neurot.2025.e00575. Online ahead of print.

ABSTRACT

Neuroblastoma is an aggressive childhood cancer that arises from the sympathetic nervous system. Despite advances in treatment, high-risk neuroblastoma remains difficult to manage due to its heterogeneous nature and frequent development of drug resistance. Drug repurposing guided by single-cell analysis presents a promising strategy for identifying new therapeutic options. Here, we aim to characterize high-risk neuroblastoma subpopulations and identify effective repurposed drugs for targeted treatment. We performed single-cell transcriptomic analysis of neuroblastoma samples, integrating bulk RNA-seq data deconvolution with clinical outcomes to define distinct malignant cell states. Using a systematic drug repurposing pipeline, we identified and validated potential therapeutic agents targeting specific high-risk neuroblastoma subpopulations. Single-cell analysis revealed 17 transcriptionally distinct neuroblastoma subpopulations. Survival analysis identified a highly aggressive subpopulation characterized by elevated UBE2C/PTTG1 expression and poor patient outcomes, distinct from a less aggressive subpopulation with favorable prognosis. Drug repurposing screening identified Adavosertib as particularly effective against the aggressive subpopulation, validated using SK-N-DZ cells as a representative model. Mechanistically, Adavosertib suppressed cell proliferation through AKT/mTOR pathway disruption, induced G2/M phase cell cycle arrest, and promoted apoptosis. Further analysis revealed UBE2C and PTTG1 as key molecular drivers of drug resistance, where their overexpression enhanced proliferation, Adavosertib resistance, and cell migration. This study establishes a single-cell-based drug repurposing strategy for high-risk neuroblastoma treatment. Our approach successfully identified Adavosertib as a promising repurposed therapeutic agent for targeting specific high-risk neuroblastoma subpopulations, providing a framework for developing more effective personalized treatment strategies.

PMID:40118716 | DOI:10.1016/j.neurot.2025.e00575

Eur J Med Chem. 2025 Mar 15;290:117516. doi: 10.1016/j.ejmech.2025.117516. Online ahead of print.

ABSTRACT

Ferroptosis has emerged as a promising therapeutic approach for a wide range of diseases. However, limited chemical diversity and poor drug-like profiles have hindered the development of effective ferroptosis inhibitors for clinical use. Herein, we identified drug-like imidazo[1,2-a]pyridine-3-amine derivatives as innovative ferroptosis inhibitors for injury-related diseases by drug scaffold repositioning strategy. Our findings established that the selected compounds exhibited high radical scavenging and effective membrane retention, thereby leading to significant suppression of lipid peroxidation and ferroptosis at nanomolar concentrations. Notably, compound C18, with low cytotoxicity and favorable pharmacokinetics properties, demonstrated remarkable in vivo neuroprotection against ischemic brain injury in mice. In conclusion, our investigations not only engender potent ferroptosis inhibitors with novel structural characteristics that warrant further development, but also serve as a valuable case study for drug repurposing in the discovery of additional ferroptosis inhibitors.

PMID:40117856 | DOI:10.1016/j.ejmech.2025.117516

Parkinsonism Relat Disord. 2025 Mar 14:107774. doi: 10.1016/j.parkreldis.2025.107774. Online ahead of print.

NO ABSTRACT

PMID:40118710 | DOI:10.1016/j.parkreldis.2025.107774

Am J Vet Res. 2025 Mar 21:1-9. doi: 10.2460/ajvr.24.12.0377. Online ahead of print.

ABSTRACT

OBJECTIVE: To determine the population variability in dextromethorphan metabolism by cytochrome (CY) P450 2D15 (CYP2D15) in dogs.

METHODS: Healthy pet dogs were recruited from 2018 through 2024 from the Inland Pacific Northwest and phenotyped by orally administering the Program in Individualized Medicine cocktail, which contains dextromethorphan, a CYP2D15-specific probe drug. Glucuronidase-treated urine samples collected 6 hours after dosing were assayed for dextromethorphan and dextrorphan concentrations. Log-transformed metabolic ratios of dextrorphan divided by dextromethorphan (DOR/DXM Log MRs) were calculated. Dogs were genotyped for 5 missense CYP2D15 variants. Univariate and multivariate statistical approaches were used to evaluate associations between DOR/DXM Log MRs and demographic variables.

RESULTS: 105 dogs, including 34 mixed breeds and 71 dogs from 20 different owner-designated breeds, were enrolled and completed the study. There was a wide distribution of DOR/DXM Log MRs, from 0.97 to 2.76, representing a log unit range of 1.8 (63-fold variation DOR/DXM Log MRs). Log-transformed metabolic ratios of dextrorphan divided by dextromethorphan were normally distributed and unimodal. The mean (± SD) DOR/DXM Log MR was 2.04 ± 0.37. Multiple linear regression analysis showed significant association (R2 = 0.16) between DOR/DXM Log MRs and dog breed for Golden Retrievers (2.26 ± 0.29; N = 23) and Pugs (1.47 ± 0.29; N = 3). Log-transformed metabolic ratios of dextrorphan divided by dextromethorphan were not associated with dog sex, age, weight, or genotype.

CONCLUSIONS: There is substantial variability in DOR/DXM Log MR values among individuals, which can be partially attributed to differences between breeds.

CLINICAL RELEVANCE: These findings predict high variability in the metabolism of drugs by CYP2D15 associated with differences between dog breeds.

PMID:40118021 | DOI:10.2460/ajvr.24.12.0377

J Cyst Fibros. 2025 Mar 20:S1569-1993(25)00074-8. doi: 10.1016/j.jcf.2025.03.006. Online ahead of print.

ABSTRACT

We present a case of a carrier mother treated with elexacaftor/tezacaftor/ivacaftor (ETI) for in-utero management of meconium ileus in a fetus diagnosed with cystic fibrosis (CF), homozygous for the F508del variant. Following multidisciplinary discussion and shared decision-making involving the parents, ETI was initiated at 27 weeks of gestation. At 38+4 weeks, the infant was delivered. Despite the treatment, the newborn developed meconium ileus, necessitating emergency surgery after birth. We explore potential factors contributing to the lack of success in our case compared to previously reported successful cases in USA and Spain. Drug levels measured in neonatal blood and in maternal breast milk indicated minimal drug exposure, raising questions about whether variability in placental transfer and excretion in breast milk or suboptimal ETI dosing in the overweight mother impacted the outcome. Additionally, the natural variability in meconium ileus outcome, which can range from spontaneous resolution to severe complications must be considered. In our case, ETI may have mitigated the severity of the condition, preventing serious complications like bowel perforation or peritonitis. However, given that about 20 % of all fetal bowel dilation resolves spontaneously, it remains uncertain whether the positive outcomes in prior cases were attributable to ETI or the natural course of the disease. We emphasize the need for more evidence on in utero ETI exposure by advocating for the collection of cases involving ETI treatment for fetal meconium ileus, regardless of outcomes. Developing guidelines will be essential to optimize benefits for both mother and fetus while minimizing risks.

PMID:40118755 | DOI:10.1016/j.jcf.2025.03.006

Radiother Oncol. 2025 Mar 19:110852. doi: 10.1016/j.radonc.2025.110852. Online ahead of print.

ABSTRACT

BACKGROUND: In the HECKTOR 2022 challenge set [1], several state-of-the-art (SOTA, achieving best performance) deep learning models were introduced for predicting recurrence-free period (RFP) in head and neck cancer patients using PET and CT images.

PURPOSE: This study investigates whether a conventional DenseNet architecture, with optimized numbers of layers and image-fusion strategies, could achieve comparable performance as SOTA models.

METHODS: The HECKTOR 2022 dataset comprises 489 oropharyngeal cancer (OPC) patients from seven distinct centers. It was randomly divided into a training set (n = 369) and an independent test set (n = 120). Furthermore, an additional dataset of 400 OPC patients, who underwent chemo(radiotherapy) at our center, was employed for external testing. Each patients' data included pre-treatment CT- and PET-scans, manually generated GTV (Gross tumour volume) contours for primary tumors and lymph nodes, and RFP information. The present study compared the performance of DenseNet against three SOTA models developed on the HECKTOR 2022 dataset.

RESULTS: When inputting CT, PET and GTV using the early fusion (considering them as different channels of input) approach, DenseNet81 (with 81 layers) obtained an internal test C-index of 0.69, a performance metric comparable with SOTA models. Notably, the removal of GTV from the input data yielded the same internal test C-index of 0.69 while improving the external test C-index from 0.59 to 0.63. Furthermore, compared to PET-only models, when utilizing the late fusion (concatenation of extracted features) with CT and PET, DenseNet81 demonstrated superior C-index values of 0.68 and 0.66 in both internal and external test sets, while using early fusion was better in only the internal test set.

CONCLUSIONS: The basic DenseNet architecture with 81 layers demonstrated a predictive performance on par with SOTA models featuring more intricate architectures in the internal test set, and better performance in the external test. The late fusion of CT and PET imaging data yielded superior performance in the external test.

PMID:40118186 | DOI:10.1016/j.radonc.2025.110852

Food Chem Toxicol. 2025 Mar 19:115401. doi: 10.1016/j.fct.2025.115401. Online ahead of print.

ABSTRACT

Acrylamide (AA), a food hazard generated during thermal processing, poses significant safety risks due to its toxicity. Conventional methods for AA toxicology are time-consuming and inadequate for analyzing cellular morphology. This study developed a novel approach combining deep learning models (U-Net and ResNet34) with cell fluorescence imaging. U-Net was used for cell segmentation, generating a single-cell dataset, while ResNet34 trained the dataset over 200 epochs, achieving an 80% validation accuracy. This method predicts AA concentration ranges by matching cell fluorescence features with the dataset and analyzes cellular phenotypic changes under AA exposure using k-means clustering and CellProfiler. The approach overcomes the limitations of traditional toxicological methods, offering a direct link between cell phenotypes and hazard toxicology. It provides a high-throughput, accurate solution to evaluate AA toxicology and refines the understanding of its cellular impacts.

PMID:40118138 | DOI:10.1016/j.fct.2025.115401

Cell Rep Med. 2025 Mar 14:102032. doi: 10.1016/j.xcrm.2025.102032. Online ahead of print.

ABSTRACT

Liver tumors, whether primary or metastatic, significantly impact the outcomes of patients with cancer. Accurate identification and quantification are crucial for effective patient management, including precise diagnosis, prognosis, and therapy evaluation. We present SALSA (system for automatic liver tumor segmentation and detection), a fully automated tool for liver tumor detection and delineation. Developed on 1,598 computed tomography (CT) scans and 4,908 liver tumors, SALSA demonstrates superior accuracy in tumor identification and volume quantification, outperforming state-of-the-art models and inter-reader agreement among expert radiologists. SALSA achieves a patient-wise detection precision of 99.65%, and 81.72% at lesion level, in the external validation cohorts. Additionally, it exhibits good overlap, achieving a dice similarity coefficient (DSC) of 0.760, outperforming both state-of-the-art and the inter-radiologist assessment. SALSA's automatic quantification of tumor volume proves to have prognostic value across various solid tumors (p = 0.028). SALSA's robust capabilities position it as a potential medical device for automatic cancer detection, staging, and response evaluation.

PMID:40118052 | DOI:10.1016/j.xcrm.2025.102032

Neural Netw. 2025 Mar 15;187:107414. doi: 10.1016/j.neunet.2025.107414. Online ahead of print.

ABSTRACT

The identification of early micro-lesions and adjacent blood vessels in CT scans plays a pivotal role in the clinical diagnosis of pancreatic cancer, considering its aggressive nature and high fatality rate. Despite the widespread application of deep learning methods for this task, several challenges persist: (1) the complex background environment in abdominal CT scans complicates the accurate localization of potential micro-tumors; (2) the subtle contrast between micro-lesions within pancreatic tissue and the surrounding tissues makes it challenging for models to capture these features accurately; and (3) tumors that invade adjacent blood vessels pose significant barriers to surgical procedures. To address these challenges, we propose LUNETR (Language-Infused UNETR), an advanced multimodal encoder model that combines textual and image information for precise medical image segmentation. The integration of an autoencoding language model with cross-attention enabling our model to effectively leverage semantic associations between textual and image data, thereby facilitating precise localization of potential pancreatic micro-tumors. Additionally, we designed a Multi-scale Aggregation Attention (MSAA) module to comprehensively capture both spatial and channel characteristics of global multi-scale image data, enhancing the model's capacity to extract features from micro-lesions embedded within pancreatic tissue. Furthermore, in order to facilitate precise segmentation of pancreatic tumors and nearby blood vessels and address the scarcity of multimodal medical datasets, we collaborated with Zhuzhou Central Hospital to construct a multimodal dataset comprising CT images and corresponding pathology reports from 135 pancreatic cancer patients. Our experimental results surpass current state-of-the-art models, with the incorporation of the semantic encoder improving the average Dice score for pancreatic tumor segmentation by 2.23 %. For the Medical Segmentation Decathlon (MSD) liver and lung cancer datasets, our model achieved an average Dice score improvement of 4.31 % and 3.67 %, respectively, demonstrating the efficacy of the LUNETR.

PMID:40117980 | DOI:10.1016/j.neunet.2025.107414

Mar Pollut Bull. 2025 Mar 19;215:117813. doi: 10.1016/j.marpolbul.2025.117813. Online ahead of print.

ABSTRACT

Efficiently monitoring red tide via satellite remote sensing is pivotal for marine disaster monitoring and ecological early warning systems. Traditional remote sensing methods for monitoring red tide typically rely on ocean colour sensors with low spatial resolution and high spectral resolution, making it difficult to monitor small events and detailed distribution of red tide. Furthermore, traditional methods are not applicable to satellite sensors with medium to high spatial resolution and low spectral resolution, significantly limiting the ability to detect red tide outbreaks in their early stages. Therefore, this study proposes a Residual Neural Network Red Tide Monitoring Model based on Spectral Information Channel Constraints (SIC-RTNet) using HY-1C/D CZI satellite data. SIC-RTNet improves monitoring accuracy through adding three key steps compared to basic deep learning methods. First, the SIC-RTNet introduces residual blocks to enhance the effective retention and transmission of weak surface signal features of red tides. Second, three spectral information channels are calculated using the four wideband channels of the images to amplify the spectral differences between red tide and seawater. Finally, an improved loss function is employed to address the issue of sample imbalance between red tides and seawater. Compared to other models, SIC-RTNet demonstrates superior performance, achieving precision and recall rates of 85.5 % and 95.4 % respectively. The F1-Score is 0.90, and the Mean Intersection over Union (MoU) is 0.90. The results indicate that the SIC-RTNet can automatically identify red tides using high spatial resolution and wideband remote sensing data, which can help the monitoring of marine ecological disasters.

PMID:40117936 | DOI:10.1016/j.marpolbul.2025.117813

Food Chem. 2025 Mar 17;480:143854. doi: 10.1016/j.foodchem.2025.143854. Online ahead of print.

ABSTRACT

Cereals are a primary source of sustenance for humanity. Monitoring, controlling, and preventing mycotoxins in cereals are vital for ensuring the safety of the cereals and their derived products. This study introduces transfer learning strategies into chemometrics to improve deep learning models applied to spectral data from different grains or toxins. Three transfer learning methods were explored for their potential to quantitatively detect fungal toxins in cereals. The feasibility of transfer learning was demonstrated by predicting wheat zearalenone (ZEN) and peanut aflatoxin B1 (AFB1) sample sets on different instruments. The results indicated that the second transfer method is effective in detecting toxins. For FT-NIR spectrometry, the transfer model achieved an R2 of 0.9356, a relative prediction deviation (RPD) of 3.9497 for wheat ZEN prediction, and an R2 of 0.9419 with an RPD of 4.1551 for peanut AFB1 detection. With NIR spectrometry, effective peanut AFB1 detection was also achieved, yielding an R2 of 0.9386 and an RPD of 4.0434 in the prediction set. These results suggest that the proposed transfer learning approach can successfully update a source domain model into one that is suitable for tasks in the target domain. This study provides a viable solution to the problem of poor adaptability of single-source models, presenting a more universally applicable method for spectral detection of fungal toxins in cereals.

PMID:40117813 | DOI:10.1016/j.foodchem.2025.143854

Comput Biol Med. 2025 Mar 20;190:110007. doi: 10.1016/j.compbiomed.2025.110007. Online ahead of print.

ABSTRACT

Skin diseases affect over a third of the global population, yet their impact is often underestimated. Automating the classification of these diseases is essential for supporting timely and accurate diagnoses. This study leverages Vision Transformers, Swin Transformers, and DinoV2, introducing DinoV2 for the first time in dermatology tasks. On a 31-class skin disease dataset, DinoV2 achieves state-of-the-art results with a test accuracy of 96.48 ± 0.0138% and an F1-Score of 97.27%, marking a nearly 10% improvement over existing benchmarks. The robustness of DinoV2 is further validated on the HAM10000 and Dermnet datasets, where it consistently surpasses prior models. Comparative analysis also includes ConvNeXt and other CNN architectures, underscoring the benefits of transformer models. Additionally, explainable AI techniques like GradCAM and SHAP provide global heatmaps and pixel-level correlation plots, offering detailed insights into disease localization. These complementary approaches enhance model transparency and support clinical correlations, assisting dermatologists in accurate diagnosis and treatment planning. This combination of high performance and clinical relevance highlights the potential of transformers, particularly DinoV2, in dermatological applications.

PMID:40117795 | DOI:10.1016/j.compbiomed.2025.110007

Ultrasonics. 2025 Mar 14;152:107641. doi: 10.1016/j.ultras.2025.107641. Online ahead of print.

ABSTRACT

This study aims to explore the feasibility of a deep learning approach to correct the distortion caused by the skull, thereby developing a transcranial adaptive focusing method for safe ultrasonic treatment in opening of the blood-brain barrier (BBB). However, aberration correction often requires significant computing power and time to ensure the accuracy of phase correction. This is due to the need to solve the evolution procedure of the sound field represented by numerous discretized grids. A combined method is proposed to train the phase prediction model for correcting the phase accurately and quickly. The method comprises pre-segmentation, k-Wave simulation, and a 3D U-net-based network. We use the k-Wave toolbox to construct a nonlinear simulation environment consisting of a 256-element phased array, a small piece of skull, and water. The skull sound speed sample combining with the phase delay serves as input for the model training. The focus volume and grating lobe level obtained by the proposed approach were the closest to those obtained by the time reversal method in all relevant approaches. Furthermore, the mean peak value obtained by the proposed approach was no less than 77% of that of the time reversal method. In this study, the computational cost of each sample's phase delay was no more than 0.05 s, which was 1/200th of the time reversal method. The proposed method eliminates the complexity of numerical calculation processes requiring consideration of more acoustic parameters, while circumventing the substantial computational resource demands and time-consuming challenges to traditional numerical approaches. The proposed method enables rapid, precise, and adaptive transcranial aberration correction on the 3D skull-based conditions, overcoming the potential inaccuracies in predicting the focal position or the acoustic energy distribution from 2D simulations. These results show the possibility of the proposed approach enabling near-real-time correction of skull-induced phase aberrations to achieve transcranial focus, thereby offering a novel option for treating brain diseases through temporary BBB opening.

PMID:40117699 | DOI:10.1016/j.ultras.2025.107641