Computers in biology and medicine最新文献

{"title":"Molecular landscape of endometrioid Cancer: Integrating multiomics and deep learning for personalized survival prediction","authors":"Behnaz Haji Molla Hoseyni ,&nbsp;Hossein Lanjanian ,&nbsp;Yasaman Zohrab Beigi ,&nbsp;Mahdieh Salimi ,&nbsp;Fatemeh Zare-Mirakabad ,&nbsp;Ali Masoudi-Nejad","doi":"10.1016/j.compbiomed.2025.110284","DOIUrl":"10.1016/j.compbiomed.2025.110284","url":null,"abstract":"<div><h3>Background</h3><div>The endometrioid subtype of endometrial cancer is a significant health concern for women, making it crucial to study the factors influencing patient outcomes.</div></div><div><h3>Method</h3><div>This study presents a novel survival analysis pipeline applied to multiomics data, including transcriptome, methylation, and proteome data, extracted from endometrioid samples in the TCGA-UCEC project to identify potential survival biomarkers. A major innovation in our work was the development of a deep learning autoencoder designed to capture the complex non-linear relationships between biological variables and survival outcomes. To achieve this, we defined a new loss function specifically for the autoencoder.</div></div><div><h3>Result</h3><div>The newly defined loss function can lead to extracting more survival information. The output of our pipeline includes 346 features ranked by their survival importance based on SHAP analysis, with a focus on the top 30 features. We analyzed the biological pathways enriched by these omics data and their contributions. As a result, we identified a relationship between Vitamin D, its receptor, and the Galanin receptor pathways with survival in endometrioid cancer.</div></div><div><h3>Conclusion</h3><div>This study introduces an innovative approach to survival analysis using multi-omics data and deep learning, with a greater focus on censored data to extract more survival information. It offers potential biomarkers for improved prognostic evaluation in endometrial cancer and presents pathway associations related to survival. These findings contribute to a better understanding of the progression of endometrial cancer.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110284"},"PeriodicalIF":7.0,"publicationDate":"2025-05-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143899810","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An enhanced harmonic densely connected hybrid transformer network architecture for chronic wound segmentation utilising multi-colour space tensor merging","authors":"Bill Cassidy ,&nbsp;Christian McBride ,&nbsp;Connah Kendrick ,&nbsp;Neil D. Reeves ,&nbsp;Joseph M. Pappachan ,&nbsp;Cornelius J. Fernandez ,&nbsp;Elias Chacko ,&nbsp;Raphael Brüngel ,&nbsp;Christoph M. Friedrich ,&nbsp;Metib Alotaibi ,&nbsp;Abdullah Abdulaziz AlWabel ,&nbsp;Mohammad Alderwish ,&nbsp;Kuan-Ying Lai ,&nbsp;Moi Hoon Yap","doi":"10.1016/j.compbiomed.2025.110172","DOIUrl":"10.1016/j.compbiomed.2025.110172","url":null,"abstract":"<div><div>Chronic wounds and associated complications present ever growing burdens for clinics and hospitals world wide. Venous, arterial, diabetic, and pressure wounds are becoming increasingly common globally. These conditions can result in highly debilitating repercussions for those affected, with limb amputations and increased mortality risk resulting from infection becoming more common. New methods to assist clinicians in chronic wound care are therefore vital to maintain high quality care standards. This paper presents an improved HarDNet segmentation architecture which integrates a contrast-eliminating component in the initial layers of the network to enhance feature learning. We also utilise a multi-colour space tensor merging process and adjust the harmonic shape of the convolution blocks to facilitate these additional features. We train our proposed model using wound images from light skinned patients and test the model on two test sets (one set with ground truth, and one without) comprising only darker skinned cases. Subjective ratings are obtained from clinical wound experts with intraclass correlation coefficient used to determine inter-rater reliability. For the dark skin tone test set with ground truth, when comparing the baseline results (<span><math><mrow><mi>D</mi><mi>S</mi><mi>C</mi><mo>=</mo><mn>0</mn><mo>.</mo><mn>6389</mn></mrow></math></span>, <span><math><mrow><mi>I</mi><mi>o</mi><mi>U</mi><mo>=</mo><mn>0</mn><mo>.</mo><mn>5350</mn></mrow></math></span>) with the results for the proposed model (<span><math><mrow><mi>D</mi><mi>S</mi><mi>C</mi><mo>=</mo><mn>0</mn><mo>.</mo><mn>7610</mn></mrow></math></span>, <span><math><mrow><mi>I</mi><mi>o</mi><mi>U</mi><mo>=</mo><mn>0</mn><mo>.</mo><mn>6620</mn></mrow></math></span>) we demonstrate improvements in terms of Dice similarity coefficient (<span><math><mrow><mo>+</mo><mn>0</mn><mo>.</mo><mn>1221</mn></mrow></math></span>) and intersection over union (<span><math><mrow><mo>+</mo><mn>0</mn><mo>.</mo><mn>1270</mn></mrow></math></span>). Measures from the qualitative analysis also indicate improvements in terms of high expert ratings, with improvements of <span><math><mrow><mo>&gt;</mo><mn>3</mn></mrow></math></span>% demonstrated when comparing the baseline model with the proposed model. This paper presents the first study to focus on darker skin tones for chronic wound segmentation using models trained only on wound images exhibiting lighter skin. Diabetes is highly prevalent in countries where patients have darker skin tones, highlighting the need for a greater focus on such cases. Additionally, we conduct the largest qualitative study to date for chronic wound segmentation. All source code for this study is available at: <span><span>https://github.com/mmu-dermatology-research/hardnet-cws</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110172"},"PeriodicalIF":7.0,"publicationDate":"2025-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143896130","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimizing stroke lesion segmentation: A dual-approach using Gaussian mixture models and nnU-Net","authors":"Adrian Mannel ,&nbsp;Dhaval Khunt ,&nbsp;Vaibhav Agrawal ,&nbsp;Kristin Schelling ,&nbsp;Eduardo Calderón ,&nbsp;Christian la Fougère ,&nbsp;Salvador Castaneda-Vega","doi":"10.1016/j.compbiomed.2025.110221","DOIUrl":"10.1016/j.compbiomed.2025.110221","url":null,"abstract":"<div><div>Machine learning-based stroke lesion segmentation models are widely used in biomedical imaging, but their ability to detect treatment effects remains largely unexplored. Gaussian Mixture Models (GMM) and nnU-Net are among the most prominent and well-established segmentation workflows. GMM has been widely used for probabilistic tissue classification for decades, while nnU-Net has established itself as a leading deep learning framework for biomedical image segmentation, with hundreds of applications in preclinical and clinical research. Despite their widespread adoption, these methods are typically evaluated using segmentation metrics alone, without assessing their reliability in detecting therapy-induced changes - a critical factor for translational research and clinical decision-making.</div><div>In this study, we systematically evaluate GMM and nnU-Net to determine their effectiveness in identifying therapy-related changes in stroke volume. Both methods demonstrate strong segmentation performance; however, nnU-Net trained solely on manual segmentations fails to detect significant therapy-induced stroke volume reductions, leading to false negative study outcomes despite achieving excellent segmentation metrics. This limitation is particularly relevant given the increasing integration of nnU-Net into biomedical research, multi-center trials and clinical workflows.</div><div>To further investigate this issue, we evaluated nnU-Net trained with GMM-derived ground truth (GT) labels and observed that it more accurately detected therapy response compared to training with Manual-GT. These results illustrate how different GT definitions can influence model performance in therapy assessment. While the integration of probabilistic methods with deep learning has been previously described, our results demonstrate its practical impact in a controlled experimental setting. By systematically evaluating two widely used segmentation methods under therapy conditions, this study highlights the importance of considering therapy detection as a key evaluation criterion, rather than relying solely on segmentation accuracy.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110221"},"PeriodicalIF":7.0,"publicationDate":"2025-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143899808","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Fast automated creation of digital twins for virtual mechanical testing of ovine fractured tibiae","authors":"Alireza Ariyanfar ,&nbsp;Mehran Bahrami ,&nbsp;Karina Klein ,&nbsp;Brigitte von Rechenberg ,&nbsp;Salim Darwiche ,&nbsp;Hannah L. Dailey","doi":"10.1016/j.compbiomed.2025.110268","DOIUrl":"10.1016/j.compbiomed.2025.110268","url":null,"abstract":"<div><div>Virtual mechanical testing with image-based digital twins enables subject-specific insights about the mechanical progression of bone fracture healing directly from imaging data. However, this technique is currently limited by the need for commercial software packages that require manual input to create finite element (FE) models from computed tomography (CT) scans. The purpose of this study was to develop automated image analysis algorithms that can create subject-specific models from CT scans without a human in the loop. Two competing techniques were developed and tested on an imaging dataset consisting of 26 intact and 44 osteotomized ovine tibiae. In both techniques, the raw image was cropped to an efficient bounding box, downsampled, segmented by an element-formation threshold, and cleaned up for efficient FE analysis using voxel-based meshes. The key difference between contour-free (CFT) and snake-reliant (SRT) techniques was threshold- and contour-based segmentation of images, respectively, before bounding box detection. The contours were detected using a snake that balanced desired aspects of the contours through energy minimization. Virtual torsion tests were performed and the results were validated by comparison to ground-truth experimental data. The CFT and SRT models produced nearly identical predictions of virtual torsional rigidity and both methods reliably replicated the physical tests. Models generated by SRT were faster to solve, but model preparation and solution combined was faster by CFT. Automatic digital twin creation by CFT is therefore recommended except where other downstream analyses require systematic spatial data sampling of the bone, which is only achieved by SRT.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110268"},"PeriodicalIF":7.0,"publicationDate":"2025-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143899809","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Balancing privacy and health integrity: A novel framework for ECG signal analysis in immersive environments","authors":"Vithurabiman Senthuran ,&nbsp;Uthayasanker Thayasivam ,&nbsp;Iynkaran Natgunanathan ,&nbsp;Keshav Sood ,&nbsp;Yong Xiang","doi":"10.1016/j.compbiomed.2025.110234","DOIUrl":"10.1016/j.compbiomed.2025.110234","url":null,"abstract":"<div><div>The widespread use of immersive technologies such as Virtual Reality, Mixed Reality, and Augmented Reality has led to the continuous collection and streaming of vast amounts of sensitive biometric data. Among the biometric signals collected, ECG (electrocardiogram) stands out given its critical role in healthcare, particularly for the diagnosis and management of cardiovascular diseases. Numerous studies have demonstrated that ECG contains traits to distinctively identify a person. As a result, the need for anonymization methods is becoming increasingly crucial to protect personal privacy while ensuring the integrity of health data for effective clinical utility. Although many anonymization methods have been proposed in the literature, there has been limited exploration into their ability to preserve data integrity while complying with stringent data protection regulations. More specifically, the utility of anonymized signal and the privacy level achieved often present a trade-off that has not been thoroughly addressed. This paper analyzes the trade-off between balancing privacy protection with the preservation of health data integrity in ECG signals focusing on memory-efficient anonymization techniques that are suitable for real-time or streaming applications and do not require heavy memory computation. Moreover, we introduce an analytical framework to evaluate the privacy preservation methods alongside health integrity, incorporating state-of-the-art disease and person identifiers. We also propose a novel metric that assists users in selecting an anonymization method based on their desired trade-off between health insights and privacy protection. The experimental results demonstrate the impact of the de-identification techniques on critical downstream tasks, such as Arrhythmia detection and Myocardial Infarction detection along with identification performance, while statistical analysis reveals the biometric nature of ECG signals. The findings highlight the limitations of using such anonymization methods and models, emphasizing the need for approaches that maintain the clinical relevance of ECG data in real-time and streaming applications, particularly in memory-constrained environments.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110234"},"PeriodicalIF":7.0,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143891337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multimodal large language models as assistance for evaluation of thyroid-associated ophthalmopathy","authors":"Bo Ram Kim ,&nbsp;Joon Yul Choi ,&nbsp;Tae Keun Yoo","doi":"10.1016/j.compbiomed.2025.110301","DOIUrl":"10.1016/j.compbiomed.2025.110301","url":null,"abstract":"<div><div>This study evaluated the potential of multimodal AI chatbots, specifically ChatGPT-4o, in assessing thyroid-associated ophthalmopathy (TAO) through the Clinical Activity Score (CAS). Using publicly available case reports and datasets, ChatGPT-4o was tasked with generating a web-based CAS calculator and estimating CAS from external ocular photographs. Its predictions were compared with CAS evaluations by ophthalmologists and convolutional neural network (CNN) models, including ResNet50. Receiver operating characteristic (ROC) areas under the curve (AUCs) were calculated for the assessment of active TAO (CAS ≥3). ChatGPT-4o demonstrated high accuracy, with mean absolute errors of 0.39 and 0.45 compared to reference ophthalmologist scores across two datasets, outperforming both Gemini Advanced and ResNet50 in identifying active TAO. In the preoperative and pre-treatment datasets, ChatGPT-4o achieved ROC-AUCs of 0.974 and 0.990, respectively, significantly exceeding the performance of ResNet50 (0.770 and 0.623). Both ChatGPT-4o and Customized GPTs achieved identical results, suggesting robust performance without the need for further customization. The AI chatbot effectively processed both text- and image-based inputs, providing detailed explanations for its CAS estimates and creating a user-friendly calculator for rapid and accessible TAO evaluation. ChatGPT-4o thus can offer a reliable tool for TAO assessment, outperforming traditional CNN-based models. Its ability to generate a CAS calculator without prior training or coding expertise highlights its practical utility for clinical ophthalmology. This study's limitations included a small sample size, lack of real-world validation, reliance on photos without patient metadata, and challenges in repeatability. Future studies should aim to validate its effectiveness in real-world clinical settings.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110301"},"PeriodicalIF":7.0,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143891338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An explainable adaptive channel weighting-based deep convolutional neural network for classifying renal disorders in computed tomography images","authors":"G. Loganathan,&nbsp;M. Palanivelan","doi":"10.1016/j.compbiomed.2025.110220","DOIUrl":"10.1016/j.compbiomed.2025.110220","url":null,"abstract":"<div><div>Renal disorders are a significant public health concern and a cause of mortality related to renal failure. Manual diagnosis is subjective, labor-intensive, and depends on the expertise of nephrologists in renal anatomy. To improve workflow efficiency and enhance diagnosis accuracy, we propose an automated deep learning model, called EACWNet, which incorporates adaptive channel weighting-based deep convolutional neural network and explainable artificial intelligence. The proposed model categorizes renal computed tomography images into various classes, such as cyst, normal, tumor, and stone. The adaptive channel weighting module utilizes both global and local contextual insights to refine the final feature map channel weights through the integration of a scale-adaptive channel attention module in the higher convolutional blocks of the VGG-19 backbone model employed in the proposed method. The efficacy of the EACWNet model has been assessed using a publicly available renal CT images dataset, attaining an accuracy of 98.87% and demonstrating a 1.75% improvement over the backbone model. However, this model exhibits class-wise precision variation, achieving higher precision for cyst, normal, and tumor cases but lower precision for the stone class due to its inherent variability and heterogeneity. Furthermore, the model predictions have been subjected to additional analysis using the explainable artificial intelligence method such as local interpretable model-agnostic explanations, to visualize better and understand the model predictions.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110220"},"PeriodicalIF":7.0,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143896128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning meets marine biology: Optimized fused features and LIME-driven insights for automated plankton classification","authors":"Muhammad Hassan ,&nbsp;Giovanna Salbitani ,&nbsp;Simona Carfagna ,&nbsp;Javed Ali Khan","doi":"10.1016/j.compbiomed.2025.110273","DOIUrl":"10.1016/j.compbiomed.2025.110273","url":null,"abstract":"<div><div>Plankton are microorganisms that play an important role in marine food webs as primary producers in the trophic web. Traditional plankton identification methods using manual microscopy and sampling are time-consuming, labor-intensive, and prone to errors. Deep learning has improved the automation of plankton identification, but it remains challenging to achieve high accuracy and efficiency in computation with limited labeled data. In this paper, we proposed an improved plankton classification model that is more accurate and interpretable. We train two models, InceptionResNetV2 (transfer learning) and DeepPlanktonNet (from scratch), on the WHOI dataset. We utilize feature fusion to supplement feature representation, merging the outputs of both models. Feature selection is achieved through the Whale Optimization Algorithm (WOA), eliminating redundancy and making it more computationally efficient. Additionally, we also employ Local Interpretable Model-agnostic Explanations (LIME) to make the model more interpretable and gain insights into how the model makes decisions. Additionally, feature selection using WOA reduces feature space and has less inference and computational cost. Our method achieves a classification accuracy of 98.79 %, which is better than previous state-of-the-art methods. For robustness testing, we train nine machine learning classifiers on the optimized features. By significantly improving classification accuracy and speed, our method enables large-scale ecological surveys, water quality monitoring, and biodiversity studies. These advances allow researchers to and environmental scientists to automate plankton classification more reliably, supporting marine conservation and resource management.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110273"},"PeriodicalIF":7.0,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143896129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"GeneDX-PBMC: An adversarial autoencoder framework for unlocking Alzheimer's disease biomarkers using blood single-cell RNA sequencing data","authors":"Hediyeh Talebi ,&nbsp;Shokoofeh Ghiam ,&nbsp;Asiyeh Mirzaei Koli ,&nbsp;Pourya Naderi Yeganeh ,&nbsp;Changiz Eslahchi","doi":"10.1016/j.compbiomed.2025.110283","DOIUrl":"10.1016/j.compbiomed.2025.110283","url":null,"abstract":"<div><h3>Objective</h3><div>To identify blood-based biomarkers and therapeutic targets for Alzheimer's disease (AD) by leveraging single-cell RNA sequencing (scRNA-seq) data from peripheral blood mononuclear cells (PBMCs) and advanced deep learning techniques.</div></div><div><h3>Methods</h3><div>Using scRNA-seq data from PBMCs of AD patients and cognitively normal controls, we developed a deep learning framework that integrates autoencoders, classifiers, and discriminators. This approach analyzed gene expression across various immune cell types—including T cells, B cells, NK cells, and monocytes—by combining both differentially expressed genes (DEGs) and subtle genetic variations typically overlooked by conventional methods. Enrichment analyses were then conducted using Gene Ontology (GO), KEGG pathways, and protein-protein interaction (PPI) networks to assess the biological relevance of the identified genes.</div></div><div><h3>Results</h3><div>Key genes, such as <em>ZFP36L2</em>, <em>PNRC1</em>, <em>DUSP1</em>, <em>BTG1</em>, <em>YBX1</em>, and <em>CYBA</em>, were identified as significant regulators of inflammation, apoptosis, and cell proliferation. Their overexpression in peripheral immune cells was linked to neuroinflammation, a critical factor in AD progression. Additionally, an observed overlap between aging-associated and AD-related genes reinforced the interconnected nature of these processes. The deep learning model achieved high precision, recall, and F1-scores across T cells, B cells, and NK cells, while Random Forest classifiers effectively managed constraints in monocyte data.</div></div><div><h3>Conclusion</h3><div>Combining scRNA-seq with deep learning provides a powerful non-invasive strategy for the early detection of AD by identifying novel blood-based biomarkers. This integrative approach not only enhances our understanding of immune regulation and neuroinflammatory pathways in AD but also paves the way for innovative diagnostic and therapeutic strategies.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110283"},"PeriodicalIF":7.0,"publicationDate":"2025-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143888276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cancer type and survival prediction based on transcriptomic feature map","authors":"Ming Yan ,&nbsp;Zirou Dong ,&nbsp;Zhaopo Zhu ,&nbsp;Chengliang Qiao ,&nbsp;Meizhi Wang ,&nbsp;Zhixia Teng ,&nbsp;Yongqiang Xing ,&nbsp;Guojun Liu ,&nbsp;Guoqing Liu ,&nbsp;Lu Cai ,&nbsp;Hu Meng","doi":"10.1016/j.compbiomed.2025.110267","DOIUrl":"10.1016/j.compbiomed.2025.110267","url":null,"abstract":"<div><div>This study achieved cancer type and survival time prediction by transforming transcriptomic features into feature maps and employing deep learning models. Using transcriptomic data from 27 cancer types and survival data from 10 types in the TCGA database, a pan-cancer transcriptomic feature map was constructed through data cleaning, feature extraction, and visualization. Using Inception network and gated convolutional modules yielded a pan-cancer classification accuracy of 91.8 %. Additionally, by extracting 31 differential genes from different cancer feature maps, an interaction network diagram was drawn, identifying two key genes, ANXA5 and ACTB. These genes are potential biomarkers related to cancer progression, angiogenesis, metastasis, and treatment resistance. Survival prediction analysis on 10 cancer types, combined with feature maps and data amplification, cancer survival prediction accuracy reached from 0.75 to 0.91. This transcriptomic feature map provides a novel approach for cancer omics analysis, to facilitate personalized treatments and reflecting individual differences.</div></div>","PeriodicalId":10578,"journal":{"name":"Computers in biology and medicine","volume":"192 ","pages":"Article 110267"},"PeriodicalIF":7.0,"publicationDate":"2025-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143891410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}