Journal of Radiation Research and Applied Sciences最新文献

{"title":"MiR-182-5p propels colorectal cancer cell proliferation, migration and invasion via targeting SIRT1","authors":"Hongwei Fan ,&nbsp;Yingxin Wang ,&nbsp;Shuqiao Zhao ,&nbsp;Rong Ai ,&nbsp;Chao Dong","doi":"10.1016/j.jrras.2025.101850","DOIUrl":"10.1016/j.jrras.2025.101850","url":null,"abstract":"<div><h3>Background</h3><div>Colorectal carcinoma (CRC) is ranked among the most malignant malignancies. MicroRNAs (miRNA) are small non-coding RNAs involved in numerous physiological or pathological processes. Abnormal microRNA-182-5p (miR-182-5p) expression has been indicated in various carcinomas. But expression profile and biology functions of miR-182-5p in CRC is controversial.</div></div><div><h3>Methods</h3><div>Here, we detected miR-182-5p level utilizing quantitative real-time PCR (qPCR), and the differences in CRC patients’ survival rates were measured utilizing log-rank test. We detected CRC cell proliferation utilizing Colony formation assay, while CRC cell migration and invasion utilizing scratch assay and transwell assay. We identified regulatory effect of miR-182-5p on sirtuin 1 (SIRT1) via Dual-luciferase reporter (DLR) assay.</div></div><div><h3>Results</h3><div>Our research indicates that miR-182-5p level is dramatically increased in CRC. For overall survival (OS) of CRC patients, miR-182-5p level acts as an independent predictor of prognosis. Moreover, miR-182-5p facilitates CRC cell proliferation, migration and invasion. Mechanistically, miR-182-5p targets SIRT1 to mediate its oncogenic role in CRC.</div></div><div><h3>Conclusion</h3><div>Therefore, the present results reveal that elevated miR-182-5p level is an independent prognostic factor for CRC patients’ OS, and miR-182-5p enhances CRC cell proliferation, migration and invasion via targeting SIRT1. Such outcomes might provide CRC with diagnostic marker and treatment target point.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 4","pages":"Article 101850"},"PeriodicalIF":2.5,"publicationDate":"2025-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144827554","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Global trends and hotspots of treatment for radiodermatitis: A bibliometric and visualization analysis (2014–2024)","authors":"Guibiao Zhang ,&nbsp;Shuilan Bao ,&nbsp;Yiren Wang ,&nbsp;Shouying Chen ,&nbsp;Hairui Deng ,&nbsp;Xiangxiang Shi ,&nbsp;Yumei Zhao","doi":"10.1016/j.jrras.2025.101864","DOIUrl":"10.1016/j.jrras.2025.101864","url":null,"abstract":"<div><div>Radiation dermatitis (RD) is a common skin toxicity of radiotherapy, affecting many patients undergoing radiation therapy. Given its impact on patient quality of life and treatment outcomes, this study utilizes advanced bibliometric tools to comprehensively analyze the research landscape related to radiation dermatitis treatment. The analysis covers various aspects, including publication trends, institutional contributions, journal coverage, and author involvement, offering a holistic view of research dynamics in this field. The study examines the evolution of treatment research across different phases of radiation dermatitis treatment by analyzing keyword co-occurrence frequencies and timeline developments. It identifies research clusters in three areas: fundamental clinical research on RD incidence, symptoms, staging, and risk factors; exploration of pathological mechanisms, including molecular and cellular bases of RD; and personalized care for healing analysis, risk identification, treatment comparison to improve prognosis and quality of life. Furthermore, the study explores the intersection of radiation dermatitis research with emerging technologies such as artificial intelligence (AI) and wearable devices, stem cell and exosome therapy, nanomaterials, and hydrogels. This integration presents unprecedented opportunities to advance radiation dermatitis care through AI-driven wound assessment, controlled exosome release via hydrogels, real-time monitoring by wearable devices, personalized treatment plans, and predictive analytics. In summary, this bibliometric study provides a comprehensive global overview of research trends, key literature, hotspots, and emerging frontiers in treatment for radiation dermatitis. Future research should continue to explore these interdisciplinary connections to develop more effective and patient-centered approaches for managing radiation dermatitis.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 4","pages":"Article 101864"},"PeriodicalIF":2.5,"publicationDate":"2025-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144827402","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Significance of radiative heat conduction on bioconvection flow of Maxwell fluid flow with homogenous-heterogenous reactions","authors":"Muhammad Naveed Khan ,&nbsp;Naveed Ahsan ,&nbsp;Awatif Alhowaity ,&nbsp;N. Ameer Ahammad ,&nbsp;Ibrahim E. Elseesy","doi":"10.1016/j.jrras.2025.101853","DOIUrl":"10.1016/j.jrras.2025.101853","url":null,"abstract":"<div><div>The study addresses the thermal radiative analysis of the bioconvective magnetized Maxwell fluid flow over a porous stretching cylinder, incorporating with variable fluid properties, homogeneous-heterogeneous reactions, and joule heating impacts. The research explores the energy transport phenomena of the fluid flow, incorporating thermal convective boundary conditions on the cylinder's surface. The flow model with energy equation is converted into the ordinary differential equations by using group of transformation. The differential equations are solved numerically using the Bvp4c technique. The study investigates the behavior of bioconvective non-Newtonian flow and heat transfer rate, focusing on temperature, velocity, concentration, and microorganism density profiles. Stronger values of the Maxwell parameter lead to a decline in fluid velocity. The study also notes that stronger resistance occurs due to larger estimates of the Darcy-Forchheimer number, resulting in reduced fluid velocity. Furthermore, the thermal conductivity parameter augments the temperature field. From the results, the study draws conclusions about the complex behavior of bioconvective Maxwell fluids under the influence of magnetohydrodynamics and chemical reactions. The work contributes to the existing literature by combining magnetohydrodynamics, chemical reactions, and bioconvection in the context of Maxwell fluids, providing insights into the interplay of these phenomena. The novelty of the work lies in its comprehensive analysis of the interrelated effects, shedding light on the behavior of viscoelastic materials under various conditions, which goes beyond previous efforts in the field.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 4","pages":"Article 101853"},"PeriodicalIF":2.5,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144827553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Bioinformatics gene analysis of potential biomarkers associated with chronic kidney disease related ischemic stroke","authors":"Mingshan Xie ,&nbsp;Ziyi Shen ,&nbsp;Guohui Jiang","doi":"10.1016/j.jrras.2025.101841","DOIUrl":"10.1016/j.jrras.2025.101841","url":null,"abstract":"<div><h3>Objectives</h3><div>Chronic kidney disease (CKD) is a key risk factor for ischemic stroke (IS), but the underlying key molecules and mechanisms linking CKD and IS remain unclear. This study aims to identify novel potential diagnostic biomarkers for chronic kidney disease-related ischemic stroke (CKD-IS) using bioinformatics and machine learning.</div></div><div><h3>Methods</h3><div>Relevant gene expression datasets (CKD: GSE37171, GSE66494; IS: GSE16561, GSE58294) were downloaded from the GEO database. Differentially expressed genes (DEGs) in peripheral blood samples from patients and healthy controls were identified, and common DEGs between CKD and IS were screened. Functional enrichment analysis (GO/KEGG) was performed to explore biological functions. Three machine learning algorithms (LASSO, Random Forest, SVM-RFE) were used to select candidate biomarkers, followed by construction of an artificial neural network model and validation via ROC curves. Immune infiltration analysis (CIBERSORT) was conducted to investigate associations between candidate genes and immune cells.</div></div><div><h3>Results</h3><div>A total of 2648 DEGs were identified in CKD datasets and 337 in IS datasets, with 29 common DEGs. GO enrichment analysis showed these common DEGs were significantly enriched in immune system processes and immune responses. Seven candidate genes (CYTM1, EVL, IFI27, PCED1B, S100A12, S100P, TMEM158) were screened by machine learning. The neural network model based on these genes showed good diagnostic performance (AUC: 0.966 in the training set), and ROC curves confirmed their diagnostic value (AUC: 0.50–0.96 in validation sets). Immune infiltration analysis revealed correlations between these genes and immune cells (e.g., neutrophils, T cells).</div></div><div><h3>Conclusion</h3><div>The seven identified candidate genes (CYTM1, EVL, IFI27, PCED1B, S100A12, S100P, TMEM158) are potential diagnostic biomarkers for CKD-IS, providing insights into the immune-related mechanisms underlying CKD-IS and supporting future precise diagnosis and treatment.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 4","pages":"Article 101841"},"PeriodicalIF":2.5,"publicationDate":"2025-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144781248","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Oestrogen inhibition of apoptosis causes upregulation of KLF5 and poor prognosis in female non-small cell lung cancer patients","authors":"Hua Liao ,&nbsp;Jiaxin Lin ,&nbsp;Miao Zhang ,&nbsp;Xingdong Cai ,&nbsp;Guangshu Chen ,&nbsp;Zewei Jiang ,&nbsp;Li Che ,&nbsp;Shengming Liu","doi":"10.1016/j.jrras.2025.101655","DOIUrl":"10.1016/j.jrras.2025.101655","url":null,"abstract":"<div><h3>Background</h3><div>KLF5 is closely related to tumour growth, proliferation, and blood vessel formation. Our study was designed to explore the role of KLF5 in non-small cell lung cancer (NSCLC) and its impact on clinical prognosis.</div></div><div><h3>Methods</h3><div>Lung tissue specimens of patients diagnosed with NSCLC were collected, and KLF5 protein expression was detected in tissues. We then compared KLF5 expression levels in NSCLC patients of different sexes and its impact on clinical features using samples from the TCGA and GEPIA databases, and performed GO and KEGG enrichment analyses on KLF5, ESR1 and related genes. Finally, in vitro experiments using NSCLC PC-9 cells demonstrated estradiol mediated regulation of KLF5 expression and apoptosis were demonstrated at the cellular level.</div></div><div><h3>Results</h3><div>KLF5 mRNA was 2.5-fold higher in NSCLC tissues vs. adjacent tissues (p = 0.003), and protein levels increased by 1.8-fold (p = 0.01), which indicated that KLF5 was significantly over-expressed in cancerous tissues of NSCLC patients. This consistent over-expression pattern was clinically consequential, as survival analysis revealed the poor prognosis of patients was related to the over-expression of KLF5. Further analysis showed that KLF5 expression was lower in female patients than in male patients. Multivariate regression analysis showed that high KLF5 expression (HR = 1.328, p &lt; 0.05) was an independent risk factor for lung cancer prognosis, but only in female patients. Oestrogen and oestrogen receptor inhibitors were shown to mediate KLF5 protein expression at the cellular level. In addition, with increasing concentrations of oestrogen, the expression of Bcl2 was progressively increased, the expression of Bax was progressively down-regulated, and the apoptosis of NSCLC cells decreases, and that this effect was dependent on the regulation of KLF5.</div></div><div><h3>Conclusion</h3><div>Estrogen upregulates KLF5, which suppresses pro-apoptotic Bax and enhances anti-apoptotic Bcl-2, reducing caspase-3 activation KLF5 may represent a potential therapeutic target for female NSCLC, warranting further preclinical and clinical investigation These findings are primarily derived from in vitro models; future in vivo studies are needed to confirm clinical relevance.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 3","pages":"Article 101655"},"PeriodicalIF":1.7,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144271143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Construction of a cancer-associated fibroblast-related lncRNA signature for forecasting the prognosis, immune infiltration and drug response of breast cancer","authors":"Kefang Chen ,&nbsp;Xiaojin Xue ,&nbsp;Weimin Yi,&nbsp;Liang Ai,&nbsp;Changzhi Yu,&nbsp;Jianjun Li,&nbsp;Qihui Huang,&nbsp;Linhui Cao","doi":"10.1016/j.jrras.2025.101685","DOIUrl":"10.1016/j.jrras.2025.101685","url":null,"abstract":"<div><h3>Objective</h3><div>Cancer-associated fibroblasts (CAFs)-associated long-non coding RNAs (lncRNAs) are key players in cancer progression. This study intended to develop a CAF-related lncRNA signature to forecast breast cancer prognosis, immune infiltration and drug response.</div></div><div><h3>Methods</h3><div>The breast cancer-related public data were downloaded. Key CAF-related lncRNAs were screened for establishing the CAF-related lncRNA prognostic signature. The clinical utility of this signature was evaluated, and a nomogram was established by combining prognostic signature with clinical factors. Furthermore, immune cell infiltration analysis, drug sensitivity analysis, and Gene Set Enrichment Analysis (GSEA) of different risk groups were conducted. Besides, the expression of signature lncRNAs was validated in five breast cancer cell lines utilizing quantitative PCR (qPCR).</div></div><div><h3>Results</h3><div>CAF levels were increased in breast cancer samples and samples with high CAF levels had poor prognosis. A prognostic signature was built by eight CAF-related prognostic lncRNAs, including LINC00844, SENCR, NR2F2-AS1, HAND2-AS1, MIR31HG, CACNA1G-AS1, LINC01120, and LINC00626, which had a high prognostic performance, with AUC value of 0.844 (0.949, 0.705) and 0.742 (0.828, 0.614) in training and validation datasets, respectively. Then, a prognostic nomogram was established, which had high survival prediction accuracy and clinically useful. The prognostic signature was related to immune cell infiltration, particularly CD8⁺ T cell and M2 macrophage, response to several drugs like AZD7762 and cisplatin, and differential pathways, including cell cycle. Besides, qPCR confirmed the reduced expression of LINC00844, SENCR, NR2F2-AS1, HAND2-AS1, MIR31HG, and CACNA1G-AS1 and elevated expression of LINC01120 and LINC00626 in breast cancer cells.</div></div><div><h3>Conclusions</h3><div>A robust CAF-related lncRNA signature was developed, offering valuable insights into forecasting prognosis and optimizing therapeutic strategies for breast cancer.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 3","pages":"Article 101685"},"PeriodicalIF":1.7,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144271144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Slip and thermal effects on MHD fractional Jeffrey nanofluid flow","authors":"Shajar Abbas ,&nbsp;S. Saleem ,&nbsp;Maryam Fatima ,&nbsp;Ioan-Lucian Popa ,&nbsp;Mudassar Nazar ,&nbsp;Ghada Ragheb Elnaggar ,&nbsp;Rashid Jan","doi":"10.1016/j.jrras.2025.101688","DOIUrl":"10.1016/j.jrras.2025.101688","url":null,"abstract":"<div><div>Nanofluids with superior thermal characteristics are increasingly important in radiation-influenced environments across industrial and biomedical domains. This work develops a fractional-order model describing the mixed convection flow of human blood-based nanofluids embedded with single-wall carbon nanotubes (SWCNTs) and multi-wall carbon nanotubes (MWCNTs) along an infinite vertical surface. To capture memory and hereditary effects, the constant proportional Caputo (CPC) operator is utilized, while the Laplace technique yields semi-analytical solutions for temperature, velocity, and concentration profiles, and its inverse Laplace transform is obtained using the Stehfest and Tzou algorithm. The results show that thermal radiation significantly enhances heat transfer, with a 12 % increase in temperature for blood-SWCNT nanofluid due to superior thermal conductivity, while blood-MWCNT nanofluid exhibits higher velocity due to lower density. These findings contribute to the development and optimization of nanofluid systems for use in industrial engineering applications, including high-performance heat exchangers, solar collectors, electronic cooling systems, and biomedical thermal therapies where precise thermal diffusion control over is essential.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 3","pages":"Article 101688"},"PeriodicalIF":1.7,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144263775","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Heat transfer and irreversibility analysis in porous vertical disk configuration with Navier slip, thermal jump and quadratic Radiation: Numerical study using Corcione's correlation","authors":"Khadija Rafique ,&nbsp;Zafar Mahmood ,&nbsp;Hamiden Abd El-Wahed Khalifa ,&nbsp;Emad A.A. Ismail ,&nbsp;Fuad A. Awwad ,&nbsp;Abhinav Kumar ,&nbsp;Seham Sh Tantawy","doi":"10.1016/j.jrras.2025.101678","DOIUrl":"10.1016/j.jrras.2025.101678","url":null,"abstract":"<div><div>This study presents the Corcione correlation, emphasizing the effects of fluid temperature, particle volume fraction, and particle size on the improvement of nanofluid heat transfer rate. This study examines the stagnation point flow of a nanofluid consisting of <span><math><mrow><msub><mrow><mi>A</mi><mi>l</mi></mrow><mn>2</mn></msub><msub><mi>O</mi><mn>3</mn></msub><mo>/</mo><msub><mi>H</mi><mn>2</mn></msub><mi>O</mi></mrow></math></span> across a vertically extending or shrinking disk. The study is conducted for nanoparticle sizes of <span><math><mrow><mn>28</mn><mo>,</mo><mn>30</mn><mtext>,</mtext></mrow></math></span> and <span><math><mrow><mn>45</mn></mrow></math></span> nm, with volume concentrations reaching up to 4 %. The Navier velocity slip and thermal slip conditions enhance the mathematical model. It also looks at the effects of magnetohydrodynamics, suction, porous media, Joule heating, quadratic thermal radiation, viscous dissipation, and the production of entropy. The bvp4c solver numerically solves the system of ordinary differential equations created by the similarity transformation of the partial differential equations. For both the stretching and shrinking scenarios, graphs and tables illustrate the effects of several relevant factors. In the event of stretching, the velocity profile exhibits a declining trend for the Navier slip parameter, porosity, magnetic field, and mass suction parameters; in the case of shrinking, however, the opposite behavior is seen. The temperature profile increases with the nanoparticle volume fraction, radiation, and temperature ratio factors in both scenarios. Both cases show an increase in entropy production for the brinkman number and a reduction in entropy generation for the temperature difference. Skin friction and Nusselt number are measured for dp = 28, 30, and 45 nm nanoparticle sizes. The Nusselt number is larger for dp = 45 nm than 28 nm and 30 nm, while skin friction is higher for 28 nm. There exists an almost perfect concordance (with a relative error of around 0 %) between the present and previous results.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 3","pages":"Article 101678"},"PeriodicalIF":1.7,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144263778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Designing a simulator for Generalized multi-hit multi-target (GMHMT) radiobiological model in cancer treatment through ionizing radiation therapy","authors":"Aryaveer Singh ,&nbsp;Rajiv Kumar Singh ,&nbsp;Fidele Maniraguha","doi":"10.1016/j.jrras.2025.101680","DOIUrl":"10.1016/j.jrras.2025.101680","url":null,"abstract":"<div><div>Radiation therapy plays a critical role in cancer treatment, necessitating accurate radiobiological models for predicting tumor cell survival. Traditional models such as the Linear-Quadratic (LQ) and Lea's target models are often inadequate in high-dose scenarios. To address this limitation, we present a MATLAB-based simulation tool implementing the Generalized Multi-Hit Multi-Target (GMHMT) model, which incorporates multi-hit and multi-target dynamics along with a negation parameter to account for repair mechanisms and non-targeted effects. This tool allows users to input tumor-specific parameters—including cell volume, number of targets and hits, and the negation factor—to simulate survival fractions under varying doses. Simulation results demonstrate strong alignment with published experimental data, achieving an RMSE of 0.1030 and an R² value of 0.9128. Sensitivity analysis reveals that dose and tumor volume are primary influencers of survival fraction, underscoring the model's relevance for personalized treatment planning. The GMHMT simulator offers improved accuracy over classical models and provides a user-friendly interface for both clinical and research applications in radiobiological modeling and therapy optimization.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 3","pages":"Article 101680"},"PeriodicalIF":1.7,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144263772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Step edge and edge loss enhanced Transformer for medical image segmentation","authors":"Juan Zheng ,&nbsp;Mei Tian ,&nbsp;Meng Zhou ,&nbsp;Jing Cai ,&nbsp;Chanzi Liu ,&nbsp;Tao Lin ,&nbsp;Haibo Si","doi":"10.1016/j.jrras.2025.101667","DOIUrl":"10.1016/j.jrras.2025.101667","url":null,"abstract":"<div><h3>Objective</h3><div>Knee osteoarthritis (KOA) is the most common degenerative joint disease, significantly impacting patients' quality of life. Currently, KOA assessments using ultrasound images rely on complex, time-consuming manual measurements. Developing automatic, accurate medical image segmentation methods for clinical diagnosis is crucial. However, due to challenges like blurred boundaries and irregular shapes in medical images, existing methods show poor edge accuracy and morphological differences.</div></div><div><h3>Methods</h3><div>To address these issues, we propose IELSEMT model. The model proposes a step edge module, incorporating it into both global and local branches during training. Additionally, after integrating the step edge module, the network's edge feature extraction capability is enhanced. During the network's backpropagation phase, we propose a weighted loss function that incorporates edge loss terms, specifically using the weighted sum of cross-entropy and mean absolute error as the network's loss function, reducing edge region errors in network segmentation results.</div></div><div><h3>Results</h3><div>Compared to the convolutional baseline, the IELSEMT model improves F1_Score by approximately 5 % and IoU by 15 %. Compared to the Med baseline model, IoU and F1 Score are about 4 % higher. Additionally, by incorporating edge loss terms during the backpropagation phase, the network training speed is enhanced. The IELSEMT model can quickly locate segmentation object edges and trains faster, achieving a convergence speed twice that of the MedT model.</div></div><div><h3>Conclusion</h3><div>The IELSEMT model proposed in this paper incorporates full-scale and edge detection concepts, introducing edge information into the decoding phase to achieve more accurate edge information acquisition, improving segmentation accuracy with clearer and more precise edges. Unlike other Transformer-based models, our proposed method does not require pre-training on large-scale datasets. Finally, we conducted experiments on knee joint ultrasound datasets, where IELSEMT demonstrated superior performance over both CNN and other related Transformer-based architectures.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"18 3","pages":"Article 101667"},"PeriodicalIF":1.7,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144263777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}