GIST Scholar
Not a member yet
30271 research outputs found
Sort by
Dysadherin/YAP axis fuels stem plasticity and immune escape in liver cancer
No full textHepatocellular carcinoma (HCC) is an aggressive malignancy that is often refractory to chemotherapy and immune checkpoint inhibitors. This therapeutic resistance is driven in part by the persistence of cancer stem-like cells (CSCs) and the development of an immune-cold tumor microenvironment. However, the upstream regulators that coordinate these malignant features remain poorly defined. In this study, we identified dysadherin as a novel upstream activator of YAP that promotes both CSC plasticity and immune evasion through the FAK/YAP/TEAD2 signaling axis. Using single-cell transcriptomic analysis, in vitro assays, and multiple in vivo models including a humanized immune mouse system, we showed that dysadherin enhances the expression of pluripotency genes, such as OCT4 and upregulates PD-L1. These changes support stem-like tumor behavior and contribute to T-cell exclusion, fostering an immunosuppressive niche. Notably, genetic knockdown or peptide-based pharmacologic inhibition of dysadherin effectively restored antitumor immune activation, suppressed metastasis and improved therapeutic responsiveness. Our findings reveal a mechanistic link between dysadherin-mediated cell adhesion signaling and the transcriptional regulation of both stemness and immune escape. Collectively, these findings establish the dysadherin/YAP axis as a key driver of HCC progression and resistance, and highlight it as a compelling therapeutic target that could overcome treatment failure in advanced liver cancer.TRUEsciescopu
Strength enhancement and analysis using laser shock peening on Inconel 738LC and silicon nitride
No full textIn this dissertation, laser shock peening (LSP) was applied on metal and ceramic material to enhance the fatigue strength and bending strength, respectively. The research consisted of two parts: an investigation of the effects of LSP on the properties, especially the fatigue strength at high temperature, of Inconel 738 low carbon (LC) (Chapter 2), and an examination of the effects of LSP on the properties of silicon nitride ceramic with sintering additives (Chapter 3).
The effectiveness of laser shock peening (LSP) at both room and elevated (850°C) temperatures were examined for Inconel 738LC, a material widely used for gas turbine blade. LSP was performed with a Nd:YAG laser (532 nm, 8 ns, 10 Hz, top-hat profile) using aluminum foil (100 μm thick) as the protective coating at the laser intensity of 10 GW/cm2 and the number of pass of 10. As the result of LSP, surface hardness increased by 38% and compressive residual stress increased about 8 times from those of the unpeened samples. Low cycle fatigue test was carried out at room temperature and elevated temperature (850°C) under strain control. For the room temperature condition, the fatigue life of laser shock peened samples increased by about 2.4 times from that of unpeened samples. At the high temperature of 850°C, the fatigue life of laser shock peened samples was overall similar to that of the unpeened samples, either slightly higher or lower depending on applied total stress level. The results showed that LSP is beneficial in enhancing fatigue strength at low temperature but its effectiveness largely disappears at a temperature as high as 850°C, possibly due to severe thermal relaxation.
Silicon nitride ceramic (Si3N4) for industrial applications is conventionally manufactured with sintering additives, and the properties of Si3N4 change significantly based on the contents of these additives. In this study, we investigate the effects of laser shock peening (LSP) on Si3N4 sintered with varying ratios of sintering additives.
The Si3N4 samples were sintered with a mixture of Y2O3, MgO, and SiO2 sintering additives at 5, 7, 9, and 11 wt%. A Nd:YAG laser (wavelength = 532 nm, maximum pulse energy = 1.4 J, repetition rate = 10 Hz, pulse duration = 8 ns, beam diameter = 11 mm, top-hat profile) was used to irradiate the Si3N4 samples. The samples were coated with a protective layer (100 μm thick aluminum foil) and a water layer to confine plasma. LSP of Si3N4 with 5% sintering additives resulted in a slight change in surface hardness but a 77% decrease in surface compressive residual stress. In contrast, LSP of Si3N4 with 11% sintering additives led to a simultaneous increase in surface hardness (7.3%) and surface compressive residual stress (67%), indicating a significant difference in the effectiveness of LSP depending on the ratio of sintering additives. Additionally, the surface of Si3N4 with 11% sintering additives showed evidence of grain refinement after LSP. It was demonstrated that the bending strength of Si3N4 with 11% sintering additives increased by 15.2%, and the depth of the fracture origin was significantly deepened.Docto
Amino Acid Hepatotoxicity Biomarkers in Human Hepatic Organoids: Promising Standardization of Drug Toxicity Evaluation
No full textHuman hepatic organoids (hHOs) are regarded as physiologically relevant in vitro platforms to evaluate hepatotoxicity, a critical step in drug development, but their applications are currently limited by the lack of qualified and standardized evaluation markers. In this study, by leveraging the established reference measurement system of amino acids (AAs), we propose 12 new biomarkers for drug-induced hepatotoxicity evaluation in human induced pluripotent stem cell-derived hHOs. Two orthogonal analytical methods for AAs were developed and validated based on isotope dilution mass spectrometry. Four AAs (aspartic acid, arginine, glutamine, and phenylalanine) and eight ratios of two designated AAs in the media of hHOs showed reliable alteration by drug treatment, which was confirmed by differentiating between hepatotoxic and nonhepatotoxic drugs. The superiorities of AA-based toxicity evaluation using the media of hHOs are as follows: (i) ability to use media only, without direct damage to or consumption of the organoids, (ii) ability to measure and compare quantities of AAs through a standardized reference measurement system rather than nonstandardized cell viability indicators, and (iii) no requirement for further data normalization in the case of the AA ratios. The AA analysis-based results demonstrate the reliability and potential of the proposed biomarkers as not only straightforward indicators of drug-induced hepatotoxicity but also absolutely comparable measures as a step toward standardization based on the AA reference measurement system. © 2025 American Chemical Society.FALSEscopu
Interfacial Polymerization of Amine-Modified SPEEK Membranes for Li+/Mg2+ Separation in Electrically Driven Systems
No full textAssociations of urine arsenic, blood manganese, and serum zinc with pterygium in Korean adults
No full textBackground Pterygium, a fleshy growth of abnormal fibrovascular tissues on eye conjunctiva, is common in older adults. The mechanism underlying pterygium development is still unknown, but oxidative stress is considered one of the main causes. Arsenic (As), which is ubiquitous in nature, may adversely affect human health by inducing oxidative stress. On the other hand, manganese (Mn) and zinc (Zn) play an important role in enhancing the antioxidant system. Objective We aimed to investigate the associations of As, Mn, and Zn with pterygium in general Korean adults using data from the Korea National Health and Nutrition Examination Survey (KNHANES) 2008–2010. Methods The study population included 2832 adults from KNHANES 2008–2009 for As and Mn analyses and 1872 adults from KNHANES 2010 for Zn analyses (the only year for which serum zinc was measured). Pterygium was diagnosed as a wing-shaped fibrovascular growth using a slit-lamp. Environmental exposure levels of total As and As species were estimated by measuring their concentrations in urine. Mn and Zn were estimated by measuring in blood and serum, respectively. Results The prevalence of pterygium was 4.9–5.6 %. After adjusting for confounding factors, the odds ratio (OR) for pterygium in the highest tertile (vs. the lowest) of total As levels in urine was 1.84 (95 % confidence interval (CI): 1.09–3.09). Total As levels had a dose-dependent association with pterygium ( p -trend = 0.021). Urinary arsenobetaine levels were further adjusted to exclude the contribution of organic As from seafood intake, and the OR for pterygium became stronger but less significant (6.54 (95 % CI: 0.82–51.92)) in the subset with As species measured (n = 280). The OR for pterygium in the second tertile of Mn levels in blood was 0.53 (95 % CI: 0.34–0.84). There was no significant association between serum Zn and pterygium. Conclusion Our findings provide epidemiological evidence that excess As and deficient Mn may be associated with pterygium in Korean adults. © 2025 Elsevier GmbH.FALSEsciescopu
Explainable AI for predicting oxidative potential of fine particles and key chemical drivers
No full textOxidative potential (OP) has emerged as promising health metric for ambient fine particles. Chemical components and OP of fine particles measured in China and Korea were used to develop OP prediction model with understanding the influence of chemical components and their interaction. Mn, Cu, Zn, Pb, and water-soluble organic carbon (WSOC) were selected as key chemical components to affect the OP. Various machine learning models incorporating explainable AI techniques were trained and evaluated. The best prediction model was found to be voting regression which aggregated individual predictions from random forest and gradient boosting models, explaining 74.9 % of OP variabilities across all measurement sites. Mn was the most important feature to affect the OP, followed by Pb, WSOC, Cu, and Zn. During OP event days at urban Gwangju, the Pb became the most important contributor, while at agricultural Gimje, the WSOC was the one to affect the OP. It was also found that the Cu above 0.004 µg/m³ with the WSOC had a strong antagonistic effect on the OP. The explainable AI methods should be so useful to predict the OP of ambient fine particles and to understand the important chemical components and their interaction. © 2025 Elsevier B.V., All rights reserved.FALSEsciescopu
Minimum EVCS Aggregation Requirements for Reliable Customer Baseline Load in Electricity Markets
No full textThe electrification of the transportation sector, driven by decarbonization policies, can significantly impact power demand patterns. Strategically integrated electric vehicles (EVs) can serve as flexible grid resources by enabling load shifting, voltage regulation, and frequency stabilization. In particular, to evaluate their potential participation in the wholesale electricity market, it is essential to assess whether aggregated electric vehicle charging stations (EVCSs) meet the eligibility and reliability criteria required for existing demand response (DR) programs. This study analyzes the reliability of aggregated EVCSs based on their aggregation level using an average-based Customer Baseline Load (CBL) method, performance metrics such as RRMSE and RMAE, and a two-stage random sampling approach. The analysis utilizes load data from 873 EVCSs located in Jeju, Korea, as of 2023. The results show that as the level of aggregation increases, both RRMSE and RMAE decrease, significantly reducing CBL estimation errors. Furthermore, it was confirmed that aggregating EVCSs beyond 10 MW of capacity or 500kWh of average hourly load achieves reliability levels comparable to those of traditional DR resources. These findings provide a practical foundation for the integration of EV-based loads into wholesale market participation and grid control strategies, while contributing to the development of CBL methods and evaluation metrics for small-scale resources. © 2025 Elsevier B.V., All rights reserved.TRUEsciescopu
CXR-LLaVA: a multimodal large language model for interpreting chest X-ray images
No full textObjectiveThis study aimed to develop an open-source multimodal large language model (CXR-LLaVA) for interpreting chest X-ray images (CXRs), leveraging recent advances in large language models (LLMs) to potentially replicate the image interpretation skills of human radiologists.Materials and methodsFor training, we collected 592,580 publicly available CXRs, of which 374,881 had labels for certain radiographic abnormalities (Dataset 1) and 217,699 provided free-text radiology reports (Dataset 2). After pre-training a vision transformer with Dataset 1, we integrated it with an LLM influenced by the LLaVA network. Then, the model was fine-tuned, primarily using Dataset 2. The model's diagnostic performance for major pathological findings was evaluated, along with the acceptability of radiologic reports by human radiologists, to gauge its potential for autonomous reporting.ResultsThe model demonstrated impressive performance in test sets, achieving an average F1 score of 0.81 for six major pathological findings in the MIMIC internal test set and 0.56 for six major pathological findings in the external test set. The model's F1 scores surpassed those of GPT-4-vision and Gemini-Pro-Vision in both test sets. In human radiologist evaluations of the external test set, the model achieved a 72.7% success rate in autonomous reporting, slightly below the 84.0% rate of ground truth reports.ConclusionThis study highlights the significant potential of multimodal LLMs for CXR interpretation, while also acknowledging the performance limitations. Despite these challenges, we believe that making our model open-source will catalyze further research, expanding its effectiveness and applicability in various clinical contexts.Key PointsQuestionHow can a multimodal large language model be adapted to interpret chest X-rays and generate radiologic reports?FindingsThe developed CXR-LLaVA model effectively detects major pathological findings in chest X-rays and generates radiologic reports with a higher accuracy compared to general-purpose models.Clinical relevanceThis study demonstrates the potential of multimodal large language models to support radiologists by autonomously generating chest X-ray reports, potentially reducing diagnostic workloads and improving radiologist efficiency.Key PointsQuestionHow can a multimodal large language model be adapted to interpret chest X-rays and generate radiologic reports?FindingsThe developed CXR-LLaVA model effectively detects major pathological findings in chest X-rays and generates radiologic reports with a higher accuracy compared to general-purpose models.Clinical relevanceThis study demonstrates the potential of multimodal large language models to support radiologists by autonomously generating chest X-ray reports, potentially reducing diagnostic workloads and improving radiologist efficiency.Key PointsQuestionHow can a multimodal large language model be adapted to interpret chest X-rays and generate radiologic reports?FindingsThe developed CXR-LLaVA model effectively detects major pathological findings in chest X-rays and generates radiologic reports with a higher accuracy compared to general-purpose models.Clinical relevanceThis study demonstrates the potential of multimodal large language models to support radiologists by autonomously generating chest X-ray reports, potentially reducing diagnostic workloads and improving radiologist efficiency.TRUEsciescopu
Real-Time Interface Prediction During Laser Processing of Thin Film Layers by High-Resolution Femtosecond Laser-Induced Breakdown Spectroscopy
No full textA novel method to determine when the laser ablation crater reaches the interface between layers in real time during the high-resolution laser processing of multilayer thin films is reported. Femtosecond laser-induced breakdown spectroscopy (LIBS; wavelength = 343nm, pulse duration = 550fs) was adopted to predict the interface location during laser ablation with lateral (~ 3μm) and depth (~ 250nm) resolutions typically required for thin-film products in the industry. Rather than identifying the intersection of the intensity profiles of the upper and lower layers, the laser shot at which the LIBS signal intensity of the lower-layer material exceeded the noise level was monitored in the proposed method. A procedure for estimating the noise level excluding non-noise peaks from the measured LIBS spectrum was introduced. It was shown that the proposed method can accurately predict when the ablation crater reaches the interface between layers by using a single LIBS spectrum even when the LIBS signal intensity fluctuates highly owing to the low pulse energy to achieve the desired spatial resolutions. Averaging of the LIBS spectra was not necessary to reduce noise, as is generally the case with noisy LIBS data. The accuracy of the proposed method was verified experimentally by examining the cross-section of an ablation crater produced with a focused ion beam. This revealed that the center of the ablation crater reached the interface, and a shallow layer near the top of the lower film was ablated at the laser shot number predicted by the proposed method. © The Author(s), under exclusive licence to Korean Society for Precision Engineering 2025.FALSEsciescopuskc