Journal of Radiation Research and Applied Sciences最新文献

{"title":"Magnetic resonance image features of children ependymoma and correlation between programmed cell death ligand 1 expression and pathological features in tumor tissues","authors":"Ling Li ,&nbsp;Mingjun Zheng ,&nbsp;Weixia Meng","doi":"10.1016/j.jrras.2026.102195","DOIUrl":"10.1016/j.jrras.2026.102195","url":null,"abstract":"<div><h3>Background and objective</h3><div>The incidence of ependymoma is higher in children than in adults, and it is a major cause of poor prognosis in pediatric patients. This study aimed to analyze the multimodal magnetic resonance imaging (MRI) characteristics of pediatric ependymoma and the correlation between programmed death-ligand 1 (PD-L1) expression in tumor tissues and clinicopathological features.</div></div><div><h3>Methods</h3><div>Fifty pediatric patients with pathologically confirmed ependymoma were enrolled. MRI, diffusion-weighted imaging (DWI), and proton magnetic resonance spectroscopy (¹H-MRS) features were evaluated according to WHO grading. PD-L1 positivity was detected in both lesional and adjacent normal tissues, and its relationship with clinicopathological characteristics and survival rate was analyzed.</div></div><div><h3>Results</h3><div>apparent diffusion coefficient (ADC)min, rADC, and NAA/Cr ratios were significantly lower in WHO grade III patients than in grades I, II, and II–III, while Cho/Cr was significantly higher (<em>P</em> &lt; 0.05). The PD-L1 positivity rate was significantly higher in tumor tissues (52 %) than in adjacent normal tissues (18 %) (<em>P</em> &lt; 0.05). PD-L1 expression was significantly elevated in patients with tumor diameter ≥2 cm, partial resection, moderate-to-poor differentiation, and WHO grade III (<em>P</em> &lt; 0.05). At the 10-month follow-up, the survival rate of PD-L1-positive patients (57.7 %) was significantly lower than that of PD-L1-negative patients (83.3 %) (<em>P</em> &lt; 0.05).</div></div><div><h3>Conclusion</h3><div>DWI can differentiate between ependymoma grades and improve preoperative diagnostic efficiency. Elevated PD-L1 expression in tumor tissue is associated with aggressive features and may serve as a marker for assessing short-term prognosis in pediatric ependymoma. MRI-DWI and ¹H-MRS parameters (ADCmin, rADC, NAA/Cr, Cho/Cr) reflect tumor grade and malignancy, while PD-L1 expression correlates with invasiveness and outcome. Their combination provides a more comprehensive basis for disease evaluation and prognostic assessment in pediatric ependymoma.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102195"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147394331","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":"A survey on recent Progress on Monte Carlo simulation methods in radiation detection","authors":"Nadyah Alanazi ,&nbsp;Raghad Aljeraiwi ,&nbsp;Maram Almutairi ,&nbsp;Abdullah N. Alodhayb","doi":"10.1016/j.jrras.2025.102146","DOIUrl":"10.1016/j.jrras.2025.102146","url":null,"abstract":"<div><div>Monte Carlo Simulations are methods used to simulate physical systems and evaluate physical properties that are extremely difficult to execute in real life. It is used as a golden standard tool for modeling complex radiation interactions with materials and for modeling complex detectors geometries. Such simulation methods designed for different particles transports such as protons, neutrons, and gamma-rays. In this review, we will focus on recent literature on using Monte Carlo Simulation in detecting radiations, focusing on GEANT, MCNP, and EGS toolkits highlighting their unique capabilities in simulating various particles and geometries with high accuracy, as well as recent advancements in enhancing their performance including high-performance computing integration, variance reduction techniques, and hybrid modeling approaches that combine Monte Carlo simulations with deterministic or AI-based techniques. The review outlines recent Monte Carlo simulation applications such as medical physics, particle physics, and environmental monitoring, underlining the adaptability and future potentials of Monte Carlo simulation in advanced radiation detection systems.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102146"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145976941","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":"Optimizing lung cancer diagnosis through computed tomography: A radiological perspective","authors":"Yazeed I. Alashban","doi":"10.1016/j.jrras.2025.102105","DOIUrl":"10.1016/j.jrras.2025.102105","url":null,"abstract":"<div><h3>Objectives</h3><div>Lung cancer remains a major global health burden with high mortality rates, primarily due to late-stage diagnosis and limitations in accessible, rapid screening techniques. This study proposes a novel deep learning framework leveraging a hybrid Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Random Forest classification (RF) mode (CNN-LSTM-RF) architecture for the automated detection and classification of lung cancer using radiological Computed Tomography (CT) scan imaging.</div></div><div><h3>Materials and methods</h3><div>For this study, DICOM (.dcm) images were obtained from The Cancer Imaging Archive (TCIA) containing a total of 355 subjects and a total of 251,135 CT scan images which was divided into 4 major classes of Adenocarcinoma, Small Cell Carcinoma, Large Cell Carcinoma, and Squamous Carcinoma. However, the study uses 249 out of 355 sample patient files containing only lung CT datasets, ignoring all other irrelevant or mixed radiology scans. Moreover, the Large Cell Carcinoma class had a very less sample size (5 patients), and thus was omitted from this study, leaving behind a total sample size of 244 patients which was considered for classification. These slices were normalized and resized to a consistent input dimension after which data augmentation (including rotations, flips, and intensity shifts) was applied to balance the dataset across classes. A custom hybrid CNN was integrated with LSTM for temporal encoding across CT slice sequences, followed by RF classification. At a distribution ratio of 70:20:10, the data was separated into training, testing, and validation datasets.</div></div><div><h3>Results</h3><div>The custom CNN-LSTM-RF model classified three cancer types – Adenocarcinoma, Small Cell Carcinoma, and Squamous Cell Carcinoma – with a validation accuracy of 97 % and an AUC exceeding 0.95 indicating a well-balanced and highly effective model. In addition to lung cancer classification, histological grading was evaluated among smoking patients, revealing that even aggressive G3 tumors do not occur with excessively higher frequency in smokers, suggesting multifactorial etiologies of lung cancer. Moreover, Early, Intermediate, and Advanced Stage T-Stage tumors were also graphically represented for different age groups. This framework demonstrates promising clinical potential for early-stage lung cancer detection, offering a scalable tool to support radiologists and reduce diagnostic delays.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102105"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145734387","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":"Botulinum toxin type A alleviates chronic post-thoracotomy pain by inhibiting the HMGB1-mediated TLR4/NF-KB signaling pathway in rats","authors":"Fuwang Wei ,&nbsp;Bingyang Lv ,&nbsp;Dan Liu ,&nbsp;Xiaoming Zou","doi":"10.1016/j.jrras.2026.102199","DOIUrl":"10.1016/j.jrras.2026.102199","url":null,"abstract":"<div><h3>Introduction</h3><div>The purpose of this study was to establish a chronic pain model after thoracotomy in rats and investigate the effects of BTX-A on spinal inflammation and the chronic hyperalgesia of rats.</div></div><div><h3>Methods</h3><div>SD rats were used to established animal model of chronic post-thoracotomy pain (CPTP) by Buvanendran method, and a control group and sham group were established. On the third day postoperatively, the rats were given subcutaneous injection of BTX-A. Pain behavior tests of rats, including Mechanical withdrawal threshold (MWT) and cold hyperalgesia threshold, were performed before operation and postoperatively. Finally, the rats were killed to collect spinal cord tissue samples for pathological examination or protein expression detection.</div></div><div><h3>Results</h3><div>After the operation, all model rats showed a decrease in MWT values and an increase in the number of positive reactions to cold acetone stimulation, the model was successfully established. In contrast to the model group, the MWT in the M + BTX-A group was significantly higher, and the number of positive reactions of cold acetone stimulation was also significantly decreased. HE staining results showed that the nucleus of spinal cord tissue was abnormally enlarged and inflammatory cells were increased in the model group. The cell morphology of M + BTX-A group was close to normal. In addition, BTX-A significantly decreased the levels of OX-42, HMGB1, TLR4, IL-10, TNF-α, p-P65 and p-IKB-α in tissues.</div></div><div><h3>Conclusion</h3><div>BTX-A was able to effectively relieve chronic post-thoracotomy pain and decrease spinal inflammation in CPTP model rats by suppressing microglial activation in spinal cord by inhibiting HMGB1-mediated TLR4/NF-κB signaling pathway.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102199"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146077889","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":"Sound-based navigation system for visually impaired individuals","authors":"Emad Malaekah ,&nbsp;Othman Alfahad ,&nbsp;Mohsen Bakouri ,&nbsp;Adel Gadallah ,&nbsp;Suvad Selman ,&nbsp;Ahmed Al Rashdi ,&nbsp;Husham Saied","doi":"10.1016/j.jrras.2026.102160","DOIUrl":"10.1016/j.jrras.2026.102160","url":null,"abstract":"<div><div>Millions of people worldwide have impaired vision for various reasons, such as cataracts, glaucoma, AMD, diabetic retinopathy, trauma, infections, genetic disorders, and amblyopia. Visual impairment ranges from normal vision to complete blindness, with varying severities that affect daily life. This study presents a sound-based navigation system to improve indoor mobility for visually impaired individuals. The system combines voice recognition, LED-guided floor paths, and real-time haptic feedback to help users follow predefined routes. Users issue voice commands to activate illuminated floor strips, which a robot cart follows using light-dependent resistor (LDR) sensors. An electromechanical compass provides accurate orientation, while a vibrating handle delivers haptic alerts for real-time directional adjustments. Bluetooth Low Energy (BLE) determines the user's location and destination to guide them along the correct route. Experimental results show high accuracy: 91.04 % voice recognition, 4 % navigation error due to sensor noise, and 3 % speed regulation error from electrical fluctuations. The proposed solution is cost-effective and user-friendly, promoting greater independence and safety for visually impaired individuals.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102160"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925211","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":"Investigation of gamma-attenuation and optical behavior in La-doped phosphate glasses for electronic shielding materials","authors":"Gharam A. Alharshan ,&nbsp;Nasra M. Ebrahem ,&nbsp;Shaaban M. Shaaban ,&nbsp;A.M.A. Mahmoud ,&nbsp;R.A. Elsad","doi":"10.1016/j.jrras.2025.102133","DOIUrl":"10.1016/j.jrras.2025.102133","url":null,"abstract":"<div><div>Phosphate glasses with a composition of 60 P₂O₅-20 Li₂O-15ZnO- 5Bi₂O₃ and (60-(<em>x</em>+0.5)) P₂O₅-20 Li₂O-15ZnO- 5Bi₂O₃- 0.5 Er₂O₃-<em>x</em>La₂O₃ with x = 0:1 mol% were formed via a quick melt-quench process in addition to casting. The produced phosphate glass samples exhibit a rise in density as well as molar volume upon substitution of P₂O₅ with a specific ratio of Er₂O₃ and La₂O₃. The UV–visible domain absorption spectra were analyzed to determine optical characteristics of the manufactured glasses. The Davis-Mott method is employed to find the Urbach energy and optical energy band gaps. The refractive index of each created glass sample is also evaluated, and these results are used to calculate the corresponding molar refraction, the metallization criterion and non-linear optical susceptibilities of all samples, using the appropriate formulae. Phy-X/PSD software was utilized to compute the mass attenuation coefficient (MAC) and linear attenuation coefficient (LAC) throughout an energy range of 0.01–15 MeV. Additionally, LAC was detected experimentally at particular energies, and there was good agreement between both experimental and theoretical findings. The accuracy of the findings was confirmed by the same outcomes that both methods produced. Furthermore, transmission factors (TF), radiation protection efficiency (RPE), (half- and tenth-value layers HVL, and TVL were estimated to evaluate the shielding effectiveness of the glass samples. As the quantity of rare earth oxides in mol% increased, radiation shielding items declared a notable enhancement, with an order of S4 &gt; S3 &gt; S2 &gt; S2 &gt; S1 found at all photon energies.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102133"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145787191","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":"A state-of-the-art statistical method with properties: Its utilization using the industrial investment funds and radiation data","authors":"Chunlian Xia ,&nbsp;Liwe Cao ,&nbsp;Maysaa Elmahi Abd Elwahab ,&nbsp;Adel M. Widyan ,&nbsp;Bassant Elkalzah","doi":"10.1016/j.jrras.2025.102085","DOIUrl":"10.1016/j.jrras.2025.102085","url":null,"abstract":"<div><div>Understanding fundamental demographic and environmental dynamics is crucial for accurate data analysis and prediction, particularly when confronting diverse geographical, economic, and physical challenges. In this context, we propose a state-of-the-art statistical model, specifically an improved family of estimators for the population distribution function, which incorporates auxiliary information to enhance estimation accuracy. This model leverages known distribution functions of auxiliary variables to construct efficient estimators for the distribution function of a target study variable. Building on established statistical techniques such as ratio, product, and classical regression estimators, our approach introduces refined alternatives that demonstrate superior statistical properties. These include reduced mean square error (MSE) and increased percentage relative efficiency (PRE) across a wide range of scenarios. Through theoretical justification and empirical validation, we show that our estimators consistently outperform traditional methods in terms of precision and robustness. To demonstrate the model’s versatility, we apply it to two complex, real-world datasets: industrial investment funds and environmental radiation data. In the financial domain, the model effectively captures trends and distributional characteristics critical to fund allocation and risk assessment. For environmental analysis, particularly radiation monitoring, the model identifies distributional shifts and anomalies, providing critical insights for ecological safety and policy-making. Our findings confirm that the proposed family of estimators constitutes a statistically powerful and adaptable modeling framework. It offers significant advancements in distribution function estimation and is well-suited for multidisciplinary applications that demand reliable inference under uncertainty. This research thus contributes to both the theory and application of modern statistical estimation in industrial and environmental sciences.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102085"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145623092","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":"KLK5 promotes anoikis resistance in non-small cell lung cancer by activating PI3K/AKT signaling pathway","authors":"Lei Zhang ,&nbsp;Qingqing Sun ,&nbsp;Qingyong Cai ,&nbsp;Mingqiang Shen ,&nbsp;Zhijun Zhu ,&nbsp;Tengfang Zhang ,&nbsp;Baolei Liang ,&nbsp;Jinjiao Zhang","doi":"10.1016/j.jrras.2026.102162","DOIUrl":"10.1016/j.jrras.2026.102162","url":null,"abstract":"<div><h3>Background</h3><div>Metastasis and recurrence remain the principal causes of mortality in non-small cell lung cancer (NSCLC) patients. A critical prerequisite for tumor cell invasion and metastasis is the development of resistance to anoikis. In NSCLC, high levels of KLK5 are linked to lymphatic metastasis and unfavorable patient survival. This study was designed to investigate the role of KLK5 in anoikis and hypothesized that KLK5 promotes NSCLC metastasis by conferring anoikis resistance via the PI3K/AKT pathway.</div></div><div><h3>Methods</h3><div>We established and characterized an anoikis-resistant stable cell line, AR-H1299. KLK5 expression in these cells was measured using RT-qPCR and Western blot. Functional assessments of AR-H1299 cells, including CCK-8 assay, colony formation assay, flow cytometry, wound healing assay, and Transwell assay, were performed, and the levels of epithelial-mesenchymal transition (EMT) and metastasis signaling-related proteins were examined. The mechanism by which KLK5 confers anoikis resistance was further investigated through Western blot analysis and functional rescue experiments.</div></div><div><h3>Results</h3><div>Our results show that KLK5 expression is closely associated with anoikis resistance in NSCLC cells. Genetic knockdown of KLK5 inhibited, while its overexpression enhanced, the proliferation, migration, invasion, and EMT process of AR-H1299 cells. Mechanistic studies revealed that KLK5 promotes the expression of downstream apoptotic regulators Bcl-2 and Bad by activating the PI3K/AKT pathway. Rescue experiments confirmed that the PI3K/AKT inhibitor LY294002 could counteract the pro-metastatic effects induced by KLK5 upregulation.</div></div><div><h3>Conclusion</h3><div>This study elucidates a previously unreported mechanism by which KLK5 confers anoikis resistance through the PI3K/AKT signaling pathway to promote metastasis in NSCLC.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102162"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145976945","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":"Impact of case management on chemotherapy adherence, complications, and quality of life post-bladder tumor resection","authors":"Qiumei Huang,&nbsp;Yudui Xia,&nbsp;Jingfen Cui,&nbsp;Xinli Wang","doi":"10.1016/j.jrras.2026.102178","DOIUrl":"10.1016/j.jrras.2026.102178","url":null,"abstract":"<div><h3>Purpose</h3><div>Successful management of non-muscle-invasive bladder cancer (NMIBC) following transurethral resection of bladder tumor (TURBT) depends critically on compliance with intravesical chemotherapy. This study evaluates the impact of a case management model on patient adherence and prognosis.</div></div><div><h3>Methods</h3><div>This retrospective cohort analysis enrolled NMIBC patients who underwent intravesical chemotherapy after TURBT from June 2022 to June 2024. Subjects were allocated to either a case management cohort (n = 47) or a conventional care cohort (n = 43). Evaluations conducted at 6 and 12 months post-intervention encompassed adherence to therapy, scores from the Functional Assessment of Cancer Therapy – Bladder (FACT-BL), incidence of complications, rates of bladder tumor recurrence, results from the Self-Rating Anxiety Scale (SAS) and Self-Rating Depression Scale (SDS), as well as outcomes from the European Organization for Research and Treatment of Cancer Quality of Life Questionnaires (EORTC QLQ-C30 and QLQ-BLS24).</div></div><div><h3>Results</h3><div>The case management group showed higher overall treatment adherence (100 % vs 83.72 %, p = 0.013), calculated for all 47 patients who completed the one-year study period with no dropouts or exclusions due to protocol deviation, and lower complication rates during instillation (21.28 % vs 46.51 %, p = 0.011). FACT-BL scores in all four domains were higher in the case management group (all p &lt; 0.05). SAS (41.77 vs 49.32, p &lt; 0.001) and SDS (44.66 vs 47.82, p = 0.001) scores were lower. QLQ-BLS24 indicated reduced future concerns and improved gastrointestinal symptoms, while QLQ-C30 showed better cognitive, emotional, and social functioning. No difference in bladder tumor recurrence rates was observed (p &gt; 0.05).</div></div><div><h3>Conclusion</h3><div>A structured case management model enhanced adherence, reduced complications, and improved quality of life for NMIBC patients receiving intravesical chemotherapy without negatively impacting short-term recurrence rates.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102178"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145976947","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":"Design of CKDNet and PMGFANet segmentation models for medical images considering accuracy improvement and data scarcity resolution","authors":"Tenghong Rong ,&nbsp;Jingqi Sun","doi":"10.1016/j.jrras.2025.102141","DOIUrl":"10.1016/j.jrras.2025.102141","url":null,"abstract":"<div><h3>Introduction/objective</h3><div>Medical image segmentation can raise the diagnostic efficiency of doctors and reduce their burden, which is of great significance. To raise the accuracy of medical image segmentation and solve the problem of scarce medical image data, corresponding segmentation algorithms are designed separately.</div></div><div><h3>Methods</h3><div>Firstly, a cascaded knowledge dissemination network model is designed to raise the accuracy of medical image segmentation through the collaborative work of three task subnets: initial segmentation, lesion classification, and fine segmentation, combined with a feature fusion module. Secondly, a semi-supervised segmentation model based on pseudo label guided feature aggregation network is designed, innovatively introducing pseudo label guided feature enhancement module and multi-scale multi-stage feature aggregation module, optimizing the teacher-student network architecture, and fully utilizing unlabeled data to supplement training samples, in order to solve the problem of data scarcity.</div></div><div><h3>Results</h3><div>The findings denoted that the average multi-category accuracy and segmentation accuracy of the cascaded knowledge dissemination network model were 86.7 % and 99.4 %. On the molecular nucleus segmentation dataset, when the proportion of annotated data was 10 % and 50 %, the maximum intersection to union ratios of the semi supervised segmentation model were 0.859 and 0.867, respectively.</div></div><div><h3>Discussion</h3><div>It can be seen that the cascaded knowledge dissemination network model has a segmentation accuracy of over 99.0 % on medical images, which can effectively improve the segmentation accuracy of medical images. The semi supervised segmentation model also has good intersection to union ratio and Dice coefficient in low annotated data ratios, and can complete segmentation tasks in medical image data scarcity. The two medical image segmentation models designed in the research have good effectiveness.</div></div><div><h3>Conclusion</h3><div>The contribution of the research is the design of a semi supervised model consisting of a cascaded knowledge dissemination network model and a pseudo label guided feature aggregation network, which improves the accuracy of medical image segmentation and solves the problem of data scarcity.</div></div>","PeriodicalId":16920,"journal":{"name":"Journal of Radiation Research and Applied Sciences","volume":"19 1","pages":"Article 102141"},"PeriodicalIF":2.5,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145837400","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}