SLAS Technology最新文献

{"title":"NLP for computational insights into nutritional impacts on colorectal cancer care","authors":"Shengnan Gong ,&nbsp;Xiaohong Jin ,&nbsp;Yujie Guo ,&nbsp;Jie Yu","doi":"10.1016/j.slast.2025.100295","DOIUrl":"10.1016/j.slast.2025.100295","url":null,"abstract":"<div><div>Colorectal cancer (CRC) is one of the most prominent cancers globally, with its incidence rising among younger adults due to improved screening practices. However, existing algorithms for CRC prediction are frequently trained on datasets that primarily reflect older persons, thus limiting their usefulness in more diverse populations. Additionally, the part of nutrition in CRC deterrence and management is gaining significant attention, although computational approaches to analyzing the impact of diet on CRC remain underdeveloped. This research introduces the Nutritional Impact on CRC Prediction Framework (NICRP-Framework), which combines Natural Language Processing (NLP) techniques with Adaptive Tunicate Swarm Optimized Large Language Models (ATSO-LLMs) to present important insights into the part of the diet in CRC care across diverse populations. The colorectal cancer dietary and lifestyle dataset, encompassing &gt;1000 participants, is collected from multiple regions and sources. The dataset includes structured and unstructured data, including textual descriptions of food ingredients. These descriptions are processed using standardization techniques, such as stop word removal, lowercasing, and punctuation elimination. Relevant terms are then extracted and visualized in a word cloud. The dataset also contained an imbalanced binary CRC outcome, which is rebalanced utilizing the random oversampling. ATSO-LLMs are employed to analyze the processed dietary data, identifying key nutritional factors and forecasting CRC and non-CRC phenotypes based on dietary patterns. The results show that combining NLP-derived features with ATSO-LLMs significantly enhances prediction accuracy (98.4 %), sensitivity (97.6 %) specificity (96.9 %) and F1-Score (96.2 %), with minimal misclassification rates. This framework represents a transformative advancement in life science by offering a new, data-driven approach to understanding the nutritional determinants of CRC, empowering healthcare professionals to make more precise predictions and adapted dietary interventions for diverse populations.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100295"},"PeriodicalIF":2.5,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143860570","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-layered data framework for enhancing postoperative outcomes and anaesthesia management through natural language processing","authors":"Peng Xu","doi":"10.1016/j.slast.2025.100294","DOIUrl":"10.1016/j.slast.2025.100294","url":null,"abstract":"<div><div>Anaesthesia management is a critical aspect of perioperative care, directly influencing postoperative recovery, pain management, and patient outcomes. Despite advancements in anaesthesia techniques, variability in patient responses and unexpected postoperative complications remain significant challenges. The research proposes a multi-layered architecture named Anaesthesia CareNet for analyzing data from diverse sources to enhance personalized anaesthesia management and postoperative outcome prediction. The architecture is structured into two primary layers: Data processing and Predictive Modeling. In the Data processing layer, advanced Natural Language Processing (NLP) techniques such as Named Entity Recognition (NER), normalization, lemmatization, and stemming are applied to clean and standardize the unstructured clinical data. Generative Pre-trained Transformer 3 (GPT-3), a Large Language Model (LLM) is employed as a feature extraction method, allowing the system to process and analyze complex clinical narratives and unstructured textual data from patient records. This enables more precise and personalized predictions, not only improving anaesthesia management but also laying the groundwork for broader applications in life sciences. The extracted data is passed into the predictive modeling layer, where the Intelligent Golden Eagle Fine-Tuned Logistic Regression (IGE-LR) model is applied. By analyzing correlations between patient characteristics, surgical details, and postoperative recovery patterns, IGE-LR enables the prediction of complications, pain management requirements, and recovery trajectories beyond anaesthesia; the methodology has potential applications in diverse areas such as diagnostics, drug discovery, and personalized medicine, where large-scale data analysis, predictive modeling, and real-time adaptability are crucial for improving patient outcomes. The proposed IGE-LR method achieves higher performance with 91.7 % accuracy, 90.6 % specificity, and 90 % AUC, with a recall of 91.3 %, precision of 90.1 %, and an F1-Score of 90.4 %. By leveraging advanced NLP and predictive analytics, Anaesthesia CareNet exemplifies how AI-driven frameworks can transform life sciences, advancing personalized healthcare and creating a more precise, efficient, and dynamic approach to treatment management.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100294"},"PeriodicalIF":2.5,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143859900","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The predictive value of chest CT combined with peripheral blood CD4/CD8 in patients with cerebral infarction complicated with pulmonary infection","authors":"Haojie Wu,&nbsp;Jijing Zheng,&nbsp;Jianhua Zhang","doi":"10.1016/j.slast.2025.100288","DOIUrl":"10.1016/j.slast.2025.100288","url":null,"abstract":"<div><div>To investigate the predictive value of chest computed tomography (CT) combined with peripheral blood CD4/CD8 in patients with cerebral infarction complicated with pulmonary infection. Lung consolidation, tree and bud sign, focus calcification ratio, C-reactive protein (CRP), procalcitonin (PCT), and interleukin-6 (IL-6) were significantly higher in the infected group than in the non-infected group, and CD4 and CD4/CD8 were significantly lower than in the non-infected group (<em>P</em> &lt; 0.05). The results of stratified regression analysis showed that CRP, PCT, IL-6, lung consolidation, tree and bud sign, and calcification all had significant negative effects on CD4/CD8 (<em>t</em>=-5.875, -3.441, -10.406, -7.741, -3.977, -6.547, all <em>P</em> &lt; 0.05). Lung consolidation, tree and bud signs, calcifications, elevated CRP, elevated PCT, and elevated IL-6 were risk factors for patients with pulmonary infection, and increased CD4/CD8 was a protective factor (<em>P</em> &lt; 0.05). There was a non-linear dose-response relationship between CD4/CD8 and the risk of concurrent pulmonary infection (<em>P</em>_{non-linearity}=0.037), with a cut-off value of 0.98. The sensitivity, specificity, positive predictive value, and negative predictive value of combined diagnosis were significantly higher than CD4/CD8 (<em>χ</em>²=6.098, 4.640, 4.643, 6.076, <em>P</em> = 0.014, 0.031, 0.031, 0.014), and the area under the ROC curve of combined diagnosis was significantly higher than chest CT and peripheral blood CD4/CD8 (<em>Z</em> = 4.018, 5.112, <em>P</em> = 0.046, 0.037). Thoracic CT combined with peripheral blood CD4/CD8 can improve the diagnostic efficiency of cerebral infarction patients complicated with pulmonary infection and provide reference for clinical diagnosis and treatment.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100288"},"PeriodicalIF":2.5,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143826452","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Nano TiO2 photocatalytic combined with optimized operating room care in postoperative infection after gynecological open abdominal surgery","authors":"Fang Yan ,&nbsp;YuXuan Qi","doi":"10.1016/j.slast.2025.100291","DOIUrl":"10.1016/j.slast.2025.100291","url":null,"abstract":"<div><h3>Background</h3><div>Disinfection of the operating room environment is essential to reduce the incidence of infection after laparotomy in obstetrics and gynecology. With the development of science and technology, nano-titanium dioxide (TiO2) photocatalytic technology has attracted much attention due to its high efficiency and low cost. To explore the effect of TiO2 photocatalytic technology and optimized operating room care in the prevention of postoperative infection in obstetrics and gynecology department.</div></div><div><h3>Methods</h3><div>Nano-TiO2 photocatalysts were prepared and characterized, and their bactericidal effect was analyzed. A total of 96 patients with gynecological open abdominal surgery were randomly divided into control group (CG) and observation group (BG), 48 cases in each group. The CG received routine care and the BG received optimized care. The wound healing rate, infection rate, serum immunoglobulin, and inflammatory factor levels were compared.</div></div><div><h3>Results</h3><div>The specific surface area of the nano-TiO2 photocatalyst was 75.1 m²/g, and the particle size was 16.6 nm, with rutile crystal structure. Compared with ultraviolet light, nano-TiO2 photocatalyst had better disinfection effect. Compared with the CG, the wound healing rate and IgG level were higher, and the infection rate, C-reactive protein, interleukin-6, and tumor necrosis factor-α were lower in the BG (<em>P</em> &lt; 0.05).</div></div><div><h3>Conclusion</h3><div>The combination of nano-TiO2 photocatalytic disinfection and optimized nursing care resulted in a 16.66 % reduction in postoperative infections and a 14.58 % improvement in wound healing. This is associated with lower airborne pathogens (66.6 CFU/m³) and improved immune-inflammatory markers (↑IgG, ↓CRP/IL-6/TNF-α).</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100291"},"PeriodicalIF":2.5,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143851488","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing Ophthalmic Anesthesia Optimization with Predictive Embedding Models","authors":"Mingdi Zhang ,&nbsp;Wanqiu Jiao ,&nbsp;Kehui Tong ,&nbsp;Ping Zhang","doi":"10.1016/j.slast.2025.100290","DOIUrl":"10.1016/j.slast.2025.100290","url":null,"abstract":"<div><div>Ophthalmic anesthesia the crucial factors in success and safety of ophthalmic surgery, which involves the delicate aspects of pain control, sedation, and patient response. Advances in ophthalmic surgery cause a need for exact and individualized anesthetic procedures to maximize patient satisfaction and outcomes. This research investigates the machine learning (ML) and natural language processing (NLP) to personalize the practice of ophthalmic anesthesia. Text data includes preoperative assessments; drug history, procedure information, and discharge summary are preprocessed using the NLP approach, stop word removal, and lemmatization. Word2Vec technique is applied for feature extraction to represent clinical terms with vectors which carry semantic meaning, helping the model comprehend the text better. This research proposes a ML algorithm of Efficient Osprey Optimized Resilient Random Forest (EOO-RRF) model to forecast ideal anesthesia plans and patient results. Experimental results show that the EOO-RRF model is superior to traditional methods and achieves metrics such as MSE = 28.424, RMSE = 4.321, AUC=98.32% and R² = 0.956. The results indicate that combining NLP and ML in ophthalmic anesthesia leads to safer, more efficient, and personalized anesthetic management.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100290"},"PeriodicalIF":2.5,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143847446","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large language models in breast cancer reconstruction: A framework for patient-specific recovery and predictive insights","authors":"Chunrao Zheng,&nbsp;Qunfang Li,&nbsp;Geling Lu,&nbsp;Yuchang Mai,&nbsp;Yuan Hu","doi":"10.1016/j.slast.2025.100285","DOIUrl":"10.1016/j.slast.2025.100285","url":null,"abstract":"<div><div>Breast cancer reconstruction, a vital part of comprehensive cancer therapy, can be performed concurrently with cancer resection, improving both physical and psychological recovery for patients. However, the intricacy and variety of recovery demand a specialized strategy. Thus, a unique framework that uses Natural Language Processing (NLP) and Large Language Models (LLMs) is developed to improve patient-specific recovery and predictive insights during breast cancer reconstruction. Lemmatization/Stemming is used for pre-processing large volumes of data from medical records, clinical notes, and treatment histories and BioBERT, a model pretrained on biomedical texts to capture complex medical terminology used for feature extraction and aids in the transformation of text data into numerical vectors. The approach employs forecasting models like ChatGPT-4 and Gemini to offer insights into the likelihood of successful reconstruction and associated problems based on specific patient characteristics, treatment options, and recovery timelines. Using sophisticated LLMs, this framework provides clinicians with a powerful tool for personalizing care by anticipating postoperative complications, recovery durations, and psychosocial consequences. Furthermore, it allows for the development of targeted rehabilitation programs that are adapted to unique patient needs, enabling greater recovery and overall quality of life. This approach not only improves clinical decision-making but also empowers patients by offering personalized recovery strategies. As a result, the accuracy of ChatGPT-4 is 98.4 % and Gemini is 98.7 %; the score per response is 2.52 for ChatGPT-4 and 2.89 for Gemini. Readability of ChatGPT-4 is 93.0 % and Gemini is 94.5 %; a relevance score is 95.5 % and 94.0 % for ChatGPT-4 and Gemini, and time response is 2.5 s for ChatGPT-4 and 2.5 s for Gemini. Finally, this research indicates how NLP and LLMs can transform breast cancer reconstruction by offering predictive insights and promoting tailored, patient-centered therapy, bridging the gap between powerful computational technologies and life science research to better patient care.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100285"},"PeriodicalIF":2.5,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143855238","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence medical image-aided diagnosis system for risk assessment of adjacent segment degeneration after lumbar fusion surgery","authors":"Bin Dai ,&nbsp;Xinyu Liang ,&nbsp;Yan Dai ,&nbsp;Xintian Ding","doi":"10.1016/j.slast.2025.100283","DOIUrl":"10.1016/j.slast.2025.100283","url":null,"abstract":"<div><div>The existing assessment of adjacent segment degeneration (ASD) risk after lumbar fusion surgery focuses on a single type of clinical information or imaging manifestations. In the early stages, it is difficult to show obvious degeneration characteristics, and the patients’ true risks cannot be fully revealed. The evaluation results based on imaging ignore the clinical symptoms and changes in quality of life of patients, limiting the understanding of the natural process of ASD and the comprehensive assessment of its risk factors, and hindering the development of effective prevention strategies. To improve the quality of postoperative management and effectively identify the characteristics of ASD, this paper studies the risk assessment of ASD after lumbar fusion surgery by combining the artificial intelligence (AI) medical image-aided diagnosis system. First, the collaborative attention mechanism is adopted to start with the extraction of single-modal features and fuse the multi-modal features of computed tomography (CT) and magnetic resonance imaging (MRI) images. Then, the similarity matrix is weighted to achieve the complementarity of multi-modal information, and the stability of feature extraction is improved through the residual network structure. Finally, the fully connected network (FCN) is combined with the multi-task learning framework to provide a more comprehensive assessment of the risk of ASD. The experimental analysis results show that compared with three advanced models, three dimensional-convolutional neural networks (3D-CNN), U-Net++, and deep residual networks (DRN), the accuracy of the model in this paper is 3.82 %, 6.17 %, and 6.68 % higher respectively; the precision is 0.56 %, 1.09 %, and 4.01 % higher respectively; the recall is 3.41 %, 4.85 %, and 5.79 % higher respectively. The conclusion shows that the AI medical image-aided diagnosis system can help to accurately identify the characteristics of ASD and effectively assess the risks after lumbar fusion surgery.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100283"},"PeriodicalIF":2.5,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143828765","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI driven cardiovascular risk prediction using NLP and Large Language Models for personalized medicine in athletes","authors":"Ang Li ,&nbsp;Yunxin Wang ,&nbsp;Hongxu Chen","doi":"10.1016/j.slast.2025.100286","DOIUrl":"10.1016/j.slast.2025.100286","url":null,"abstract":"<div><div>The performance and long-term health of athletes are significantly influenced by their cardiovascular resilience and associated risk factors. This study explores the innovative applications of Natural Language Processing (NLP) and Large Language Models (LLMs) in biomedical diagnostics, particularly for AI-driven arrhythmia detection, hypertrophic cardiomyopathy (HCM) in athletes, and personalized medicine. The complexity of analysing diverse biomedical datasets, such as electrocardiograms (ECG), clinical records, genetic screening reports, and imaging results, poses challenges in obtaining precise early diagnoses. To address these issues, we introduce a hybrid machine learning (ML) framework that integrates the Wolf Pack Search Algorithm Dynamic Random Forest (WPSA-DRF) with a RoBERTa-based LLM to enhance the accuracy of cardiovascular disease predictions. Using advanced NLP techniques, including biomedical text mining, entity recognition, and feature extraction, the system processes structured and unstructured clinical data to detect abnormalities associated with sudden cardiac arrest (SCA), arrhythmias, and genetic cardiomyopathies. The proposed system achieves a diagnostic accuracy of 92.5 %, precision of 92.7 %, recall of 99.23 %, and F1-score of 95.6 %, outperforming traditional diagnostic methodologies. Furthermore, the research underscores the role of LLMs in personalized medicine, identifying patient-specific risk factors and optimizing treatment pathways for cardiac patients. This work highlights how NLP-driven AI solutions are transforming biomedical research, accelerating early disease detection, and improving clinical decision-making for both athletes and the general population.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100286"},"PeriodicalIF":2.5,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143855237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A hybrid PKI and spiking neural network approach for enhancing security and energy efficiency in IoMT-based healthcare 5.0","authors":"Dipalee D․Rane Chaudhari ,&nbsp;Manisha S. Bhende ,&nbsp;Aadam Quraishi ,&nbsp;Azzah AlGhamdi ,&nbsp;Ismail Keshta ,&nbsp;Mukesh Soni ,&nbsp;Brajesh Kumar Singh ,&nbsp;Haewon Byeon ,&nbsp;Mohammad Shabaz","doi":"10.1016/j.slast.2025.100284","DOIUrl":"10.1016/j.slast.2025.100284","url":null,"abstract":"<div><div>In the rapidly evolving field of healthcare 5.0, the Internet of Medical Things (IoMT) is expected to be an enabler that allows smart medical devices to collaborate and communicate with healthcare networks to speed up procedures, enhance care, and improve disease management. However, one of the critical issues for these networks still remains the secure and energy-efficient transmission of sensitive patient data. Thus, a novel security framework is proposed in this work, in which a Public Key Infrastructure- Energy-Efficient Routing Protocol (PKI-EERP) with a Zebra Optimization Algorithm (ZOA) is incorporated in spiking neural networks. The method combines data security robustness of the spiking neural networks to detect anomalies and check for access control purposes, with the PKI encryption to provide safe encryption and key management. The ZOA optimizes energy consumption in WSNs, and as a result transmission energy is significantly reduced up to 35 % compared to other implementations, and the network lifetime is increased by about 30 % through effective load balancing. It enhances both the privacy and energy efficiency that are essential for the safe and reliable operation of IoMT systems in contemporary healthcare environments, thus improving patient outcomes as well as standards of operations.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100284"},"PeriodicalIF":2.5,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143851489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Association of the characteristics of brain magnetic resonance imaging with genes related to disease onset in schizophrenia patients","authors":"Jiantu Lin ,&nbsp;Bo Wang ,&nbsp;Shaoguang Chen ,&nbsp;Fengling Cao ,&nbsp;Jingbin Zhang ,&nbsp;Zirong Lu","doi":"10.1016/j.slast.2025.100281","DOIUrl":"10.1016/j.slast.2025.100281","url":null,"abstract":"<div><h3>Background</h3><div>Schizophrenia (SCH) is a complex neurodevelopmental disorder, whose pathogenesis is not fully elucidated. This article aims to reveal disease-specific brain structural and functional changes and their potential genetic basis by analyzing the characteristics of brain magnetic resonance imaging (MRI) in SCH patients and related gene expression patterns. Methods: Differentially expressed genes (DEGs) between SCH and healthy control (NC) groups in the GSE48072 dataset were identified and functionally analyzed, and a protein-protein interaction (PPI) network was fabricated to screen for core genes (CGs). Meanwhile, MRI data from the COBRE, the Human Connectome Project (HCP), the 1000 Functional Connectomes Project (FCP), and the Consortium for Reliability and Reproducibility (CoRR) were utilized to explore differences in brain activity patterns between SCH patients and NC group using a 3D deep aggregation network (3D DANet) machine learning approach. A correlation analysis was performed between the identified CGs and MRI imaging characteristics. Results: 82 DEGs were collected from the GSE48072 dataset, primarily involved in cytotoxic granules, growth factor binding, and graft-versus-host disease pathways. The construction of the PPI network revealed <em>KLRD1, KLRF1, CD244, GZMH, GZMA, GZMB, PRF1</em>, and <em>SLAMF6</em> as CGs. SCH patients exhibited relatively enhanced activity patterns in the frontoparietal attention network (FAN) and default mode network (DMN) across four datasets, while showing a trend of weakening in most other networks. The 3D DANet demonstrated higher accuracy, specificity, and sensitivity in brain image classification. The correlation between enhancement of the DMN and genetic abnormalities was the strongest, followed by the enhancement of the frontal and parietal attention networks. In contrast, the correlation between the weakening of the sensory-motor network and occipital network and genetic abnormalities was relatively weak. The strongest correlation was observed between MRI characteristics and the <em>KLRD1</em> and <em>CD244</em> genes. Conclusion: The granzyme-mediated programmed cell death signaling pathway is related to pathogenesis of SCH, and <em>CD244</em> may serve as potential biological markers for diagnosing SCH. The correlation between enhancement of the DMN and genetic abnormalities was the strongest, followed by the enhancement of the frontal and parietal attention networks. In contrast, the correlation between weakening of the sensory-motor network and occipital network and genetic abnormalities was relatively weak. Additionally, the strongest correlation was observed between MRI features and the <em>KLRD1</em> and <em>CD244</em> genes. The use of the 3D DANet method has improved the detection precision of brain structural and functional changes in SCH patients, providing a new perspective for understanding the biological basis of the disease.</div></div>","PeriodicalId":54248,"journal":{"name":"SLAS Technology","volume":"32 ","pages":"Article 100281"},"PeriodicalIF":2.5,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143755687","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}