Browse Articles
Discover research articles across all indexed journals
A customized AI-based machine learning model for evaluating participants’ activities under a workshop setting
Adaptive queue management in healthcare using supervised Q-learning with time-varying reward and cost structures
Advancing biomedical data analytics using explainable neural network-based learning model for progressive neurodegenerative disorder diagnosis
Abstract Huntington’s disease (HD) is an inherited neurological disease caused by variations in the huntingtin (HTT) gene, which leads to neuronal degeneration. Conventionally, HD is affiliated with the gathering and misfolding of mutant HTT arising from an increased number of CAG triplets. Artificial Intelligence has emerged as an important tool in healthcare, supporting the monitoring, detection, and management of HD. Machine learning and deep learning methods are widely used for automated HD identification using neuroimaging, genetic, and clinical data. However, most DL models behave like a black box, making it difficult to interpret decision-making from clinical data, which reduces trust in medical applications. Therefore, this study presents an Explainable Neural Network-Driven Learning Model for Neurodegenerative Disorder Diagnosis (XNNLM-NDD). The primary objective of the proposed model is to examine clinical attributes and identify disease patterns efficiently for precise diagnosis. The model performs feature selection using a hybrid combination of minimum redundancy maximum relevance and ReliefF methods to select the most informative and non-redundant features from the dataset. For classification, the proposed approach employs a feature tokenizer-transformer model, which can capture complex feature interactions and improve classification accuracy on structured medical data. Furthermore, the model is optimized using the Cycle-Norm-Adam algorithm. For ensuring model transparency and interpretability, SHAP-based explainable artificial intelligence method is used to highlight the contribution of each feature towards the final prediction. The experimental evaluation is carried out on the Huntington Disease Dataset sourced from Kaggle. The results show that the proposed XNNLM-NDD approach accomplishes improved performance with an accuracy of 96.50% compared to existing techniques, indicating its efficiency in progressive neurodegenerative disorder diagnosis.
Prophage genomics of carbapenemase-producing Klebsiella pneumoniae from animal-derived food sources
Perceived and received support among patients with cancer receiving palliative care in Malta: a qualitative study
MST1/Drp1 axis mediates microglia pro-inflammatory activation following cerebral ischemia-reperfusion injury
Abstract The present study investigated the role of the mammalian sterile 20-like kinase 1/dynamin-related protein 1 (MST1/Drp1) axis in regulating microglia pro-inflammatory activation during cerebral ischemia-reperfusion injury (CIRI). An in vivo model of middle cerebral artery occlusion/reperfusion (MCAO/R) in rats and an in vitro oxygen-glucose deprivation/reoxygenation (OGD/R) model in BV-2 microglial cells and primary microglia were established. Inhibitors of MST1 (XMU-MP-1) or/and Drp1 (Mdivi-1), along with genetic approaches including siMST1-mediated knockdown and plasmid-based overexpression, were utilized in the models. The expression and activation of MST1 and Drp1, mitochondrial morphology changes, microglia pro-inflammatory activation makers, pro-inflammatory cytokine release, DNA fragmentation and neurological function were evaluated. The findings indicated that reperfusion or reoxygenation led to a rise in total and phosphorylation levels of MST1 and Drp1. The reperfusion also facilitated the Drp1 translocation toward mitochondria, and resulted in increased mitochondrial morphological changes. MST1 or/and Drp1 inhibitors decreased p-MST1 and p-Drp1(Ser616) levels, attenuated mitochondrial fission, suppressed microglia pro-inflammatory activation, pro-inflammatory factors release (TNF-α, IL-6 and IL-1β). Overall, these effects ultimately mitigated cerebral injury as evidenced by reduced DNA fragmentation, decreased cerebral infarct volumes, and improved neurological function. Combined inhibitors further exerted ameliorative effects on the above-mentioned parameters. In the in vitro experiments, siMST1 knockdown attenuated p-Drp1(Ser616) expression and suppressed microglia pro-inflammatory activation under OGD/R conditions. These protective effects were reversed by Drp1 overexpression. These findings indicate that p-MST1 drives microglia pro-inflammatory activation via promoting the p-Drp1(Ser616)-mediated excessive mitochondrial fission during CIRI.
AI‑driven framework for contract risk automation and compliance in oracle CPQ
Leveraging CT-derived chronic imaging signatures for acute kidney injury evaluation
Genetic profiling enhances cystic fibrosis prenatal diagnosis
Molecular evidence for early deuterostome origins of ovarian cell types and neuroendocrine control of reproduction
Automatic pain assessment from facial action units in ICU patients via various machine learning models
Abstract Pain represents a critical vital sign monitored in intensive care unit (ICU) patients. The facial action coding system (FACS) defines facial action units (AUs) and provides a structured framework for pain recognition. Currently, the most broadly used pain assessment method is the pain intensity scale developed by Prkachin and Solomon (PSPI), which relies on predefined AUs to quantify facial expressions. However, due to the influence of underlying diseases and facial texture variations in ICU patients, AUs can fail to transfer to clinical settings accurately. To address this problem, this study uses video sequences of pain states collected from 61 ICU patients under resting, daily, and procedural conditions by using an advanced AU detection system. By evaluating the AUs with statistical features and various classification models, this study identifies six key AUs that outperform the PSPI’s predefined AUs in terms of accuracy, precision, recall, and F1-score metrics. Further, this study explores the performance of various temporal self-learning networks in the pain assessment task, thus further validating the effectiveness of the identified AU combination. The results presented in this study demonstrate that using AU dynamic learning in combination with deep temporal analysis can improve the reliability of clinical pain assessment. Finally, this study offers a promising approach for automated pain monitoring systems in ICU settings.
In situ reprogramming of CAR-alveolar macrophages via liposomal nanomedicine for lung cancer immunotherapy
Multifunctional bioactivity of Portulaca oleracea seed oil: GC-MS-based characterization and in vitro evaluation
Abstract Purslane ( Portulaca oleracea L.) is a widespread weed highly valued for its nutritional value, particularly its omega -3 fatty acid content. This study aimed to investigate the pharmacological properties and the phenolic and flavonoid contents of P. olec racea seed oil. P. oleracea seed oil was tested for antioxidant activity (DPPH· and ABTS), cell viability and cytotoxicity and anti-inflammatory activity. The contents of total phenolics, flavonoids, and constituents were characterized. The seed oil of P.oleracea had high flavonoid (4.53 mg/g) and phenolic (8.65 mg/g) content. The antioxidant activity of P. oleracea seed oil increased from 24.16% at a concentration of 0.5 µg/mL to 97.28% at a concentration of 1000 µg/mL. The MTT assay demonstrated that P. oleracea seed oil exhibited relatively weak to moderate cytotoxic activity against the A549, HepG-2, and MCF-7 cell lines. Lower concentrations of the seed oil extract exhibited anti-inflammatory activity, and these effects were compared with those of indomethacin as a positive control. GC-MS analysis showed that the seed oil of P. oleracea is rich in olean-12-ene-3,28-diol and 9-octadecenoic acid. It could be concluded that P. oleracea seed oil exhibits notable in vitro antioxidant and cytotoxic activities, which may be attributed to its content of flavonoids and phenolic compounds. However, further in vivo studies and toxicity evaluations are required to confirm its potential applications.
Characterization of membrane structures regulating primary ciliogenesis by quantitative isotropic ultrastructure imaging
Abstract The trafficking, docking, and fusion of membrane vesicles at the mother centriole (MC) are important for primary cilium construction. Here, we determined the three-dimensional (3D) membrane ultrastructures, and associated proteins, involved in this primary cilium assembly mechanism upstream of axoneme growth. Our work suggests that the enlargement of small vesicles docked to the MC is a key trigger for ciliogenesis progression, a process requiring the MC distal appendage protein CEP164. These vesicles appear to fuse to form tubular C-shaped intermediates and an unprecedented toroidal membrane intermediate. The formation of these previously uncharacterized tubular membrane ciliogenesis intermediates is orchestrated by the membrane trafficking regulators EHD1 and RAB8, and is associated with the IFT-B complex protein IFT88. Remarkably, we show that EHD1, through its membrane tubulation function, regulates ciliogenesis progression by directly promoting CP110/CEP97 removal from the MC cap. The establishment of these tubular membrane structures is also associated with the recruitment of the ciliary gate transition zone proteins. Together, these findings redefine the architectural framework of early ciliogenesis and underscore the utility of isotropic ultrastructural imaging combined with quantitative 3D analysis for elucidating mechanisms of membrane trafficking and organelle biogenesis.
Pharmacological treatments and clinical events in newly diagnosed heart failure patients stratified by ejection fraction in Japan
Abstract There is a limited understanding of the uptake of pharmacological treatments and incidence of clinical events in newly diagnosed heart failure (HF) patients, stratified by left-ventricular ejection fraction (LVEF) in contemporary clinical settings in Japan. A retrospective cohort study was conducted using a nationwide Japanese hospital database to evaluate the patterns of HF medications and clinical events in adult patients with a first confirmed HF diagnosis from January 1, 2020–July 31, 2023 (n = 16,001). Fine-Gray sub-distribution hazard models were applied to assess factors associated with HF medication initiations and clinical events. Overall, 5473 (34.2%), 3053 (19.1%), and 7475 (46.7%) patients with HF with reduced ejection fraction, mildly reduced ejection fraction (HFmrEF), and preserved ejection fraction (HFpEF) were included, respectively. Within the first 6 months, prescription rates of HF medications were: angiotensin-converting enzyme inhibitors, angiotensin-receptor blockers, or angiotensin receptor blocker-neprilysin inhibitor (65.6%), beta-blockers (57.6%), mineralocorticoid antagonists (55.7%), and sodium-glucose cotransporter-2 inhibitors (33.2%), with the lowest rates in HFpEF patients. The increase in prescription rates between 6 and 12 months was modest across all medication classes. One-year incidence rates per 100 person-years (95% confidence interval) of in-hospital all-cause mortality and in-hospital cardiovascular death were 23.0 (22.1–24.0) and 16.6 (15.8–17.4), respectively. HFmrEF and HFpEF were associated with lower hazards of treatment initiation for most HF medications, whereas the risks of in-hospital all-cause mortality and in-hospital cardiovascular death did not differ significantly across all LVEF subtypes. The findings underscore the suboptimal uptake of pharmacological treatments, despite the poor prognosis of newly diagnosed HF patients. The disparities in initiations of recommended treatments reaffirm the importance of properly implementing evidence-based therapies within each LVEF subtype.