[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-349281-105":3,"doc-detail-349281-en":73,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":66,"head_meta":68,"extra_data":70,"updated_unix":72},105,"en","skin-cancer-detection-using-late-fusion-of-pretrained-models","Skin cancer detection using late fusion of pretrained models","","Skin cancer is a prevalent and deadly malignancy worldwide, making reliable early diagnosis essential for effective clinical intervention. Existing deep learning systems often overfit and show limited generalizability across lesion types. This work proposes a late-fusion ensemble that integrates pretrained convolutional neural networks with Diverse Convolution Networks (DCNs) to form a complementary, discriminative detection framework from dermoscopic images. Experiments on 10,600 ISIC dermoscopic images achieve 99.7% accuracy and strong performance on melanoma, supporting robustness in complex clinical settings.",{"@graph":14,"@context":65},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/healthcare/","Healthcare",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/skin-cancer-detection-using-late-fusion-of-pretrained-models/349281/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/skin-cancer-detection-using-late-fusion-of-pretrained-models/349281.png","ImageObject",300,407,{"name":42,"@type":43},"Mabel","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-27","2026-09-22",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",11,{"@type":57,"mainEntity":58},"FAQPage",[59],{"name":60,"@type":61,"acceptedAnswer":62},"What future work is planned?","Question",{"text":63,"@type":64},"Future work focuses on model explainability, integration with clinical decision support systems, and validation using larger, multi-source datasets to improve real-world applicability.","Answer","https://schema.org",{"og:url":32,"og:type":67,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":69,"canonical":32},"index,follow",{"doc_id":71,"site_id":7},349281,1790102223,{"code":4,"msg":74,"data":75},"success",{"doc_id":71,"user_id":76,"nickname":42,"user_avatar":77,"doc_module":4,"category_id":78,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":79,"file_id":80,"file_url":81,"file_type":82,"file_size":83,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":84,"language":85,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":86,"faqs":87,"seo_title":88,"seo_description":12,"update_tm":89,"read_time":90},7971474920318,"https://ap-avatar.wpscdn.com/avatar/74000ee537e6b0ca360?x-image-process=image/resize,m_fixed,w_180,h_180&k=1788430171632181674",7,"[www. nature.com/scientificreports](www. nature.com/scientificreports)  \nOPEN  \nSkin cancer detection using late fusion of pretrained models  \nAmel Ksibi1􀀍, Ahlem Walha2, Mohammed Zakariah3, ManelAyadi1,  \nNouf Abdullah Almujally1 & Tagrid Alshalali1  \nSkin cancer is one of the most common and fatal malignancies worldwide, and reliable, early diagnosis systems are required to provide effective clinical intervention. Nevertheless, even with recent progress in deep learning, current models tend to overfit and lack generalizability and robustness across various lesion types. This paper overcomes these challenges by presenting a late-fusion ensemble of pretrained convolutional neural networks (CNNs) integrated with Diverse Convolution Networks (DCNs) to build a complementary and discriminative skin cancer detection framework from dermoscopic images. The study utilized a curated dataset of 10,600 dermoscopic images from the ISIC Archive, comprising 9600 images for training and 1000 images for testing, encompassing both malignant and benign lesion classes. The proposed late-fusion model combines the predictions of several pre-trained architectures, leveraging their strengths while mitigating bias. Experimental results show excellent performance with an accuracy of 99.7%, an F1-score of 99.75%, a precision of 99.12%, and a recall of 99.34% on the test set, and an accuracy of 99.1% and an F1-score exceeding state-of-the-art models of 99.8% on the melanoma dataset. The conclusions validate the robustness and diagnostic accuracy of the framework in complex clinical situations. Future work will focus on model explainability, integration with clinical decision support systems, and validation using larger, multi-source datasets to enhance real-world applicability.  \nKeywords Skin cancer detection, ISIC database, Melanoma, Late fusion of pre-trained models, Diverse convolution networks  \nAbbreviations  \nML  \nDCN  \nCNN  \nSVM  \nDCN  \nROC CapsNet Nb  \nISIC  \nKNN  \nLDA  \nRNNTP/ FPAI  \nDL  \nWHO  \nDNN  \nAUC  \nRF  \nDT  \nMSCD-Net  \nMachine-learning  \nDiverse convolution network Convolutional neural network Support vector machine  \nDeep convolutional neural network Receiver operating characteristics Capsule network  \nNaïve Bayes  \nInternational skin imaging collaboration K-nearest neighbors algorithm  \nLinear discriminant analysis Recurrent neural networks True Positive/ False Positive Artificial intelligence Deep-learning  \nWorld Health Organization Deep neural network Area under the ROC curve Random forest  \nDecision tree  \nMulti-class skin cancer detection network  \n1Department of Information Systems, College of Computer and Information Sciences, Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia. 2Department of Computer Science, College of Engineering in Al-Lith, Umm Al-Qura University, Makkah, Saudi Arabia. 3Department of Computer Science and Engineering, College of Applied Studies, King Saud University, P.O. Box 22459, Riyadh 11495, Saudi Arabia. 􀀍 email: [amelksibi@pnu.edu.sa](amelksibi@pnu.edu.sa)  \n[www. nature.com/scientificreports/](www. nature.com/scientificreports/)  \nCV MSCDNet ReLU TN/ FN EDA  \nCross-validation  \nMulti-class skin cancer detection network Rectified linear unit  \nTrue Negative/ False Negative Exploratory data analysis  \nSkin cancer has become a major health problem throughout the world, with melanoma being the deadliest form of skin cancer. The disturbing increasing trend of events has highlighted the importance of timely and accurate diagnosis1,2. Early diagnosis allows the severity of skin cancer to be minimized and the patient’s outcome tobe improved3. In this regard, analysing medical images plays a crucial role in medical research, enabling the detection of abnormalities in various body parts4. The results of the study clearly demonstrate the need to utilize complex methods for the diagnosis of skin cancer. The primary goal is to enhance the utilization of novel, diversified convolutional networks (D","cbCaimgkOPrEMBOX","https://ap.wps.com/l/cbCaimgkOPrEMBOX","pdf",5075100,33,"English","# Introduction\n## Problem and motivation\n## Skin cancer epidemiology and risks\n## Early diagnosis and medical imaging\n# Method and framework\n## Late-fusion ensemble with pretrained CNNs\n## Diverse Convolution Networks (DCNs)\n# Dataset and evaluation\n## ISIC dataset split for training and testing\n## Performance metrics and results\n# Conclusions and future work\n## Explainability and clinical decision support\n## Validation on larger multi-source datasets","[{\"question\":\"What future work is planned?\",\"answer\":\"Future work focuses on model explainability, integration with clinical decision support systems, and validation using larger, multi-source datasets to improve real-world applicability.\"}]","Skin cancer detection using late fusion of pretrained models | PDF",1790082585,83,{"code":4,"msg":74,"data":92},[93,97,101,105,110,115,118,123,128,131,135],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":25,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":25,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":78,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":116,"slug":117},40,"healthcare",{"id":119,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":121,"slug":122},8,"Research & Report",30,"research-report",{"id":124,"doc_module":4,"doc_module_name":25,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":25,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":25,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":25,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]