[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-351834-105":59,"doc-detail-351834-en":131},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","from-local-textures-to-global-attention-a-hybrid-feature-fusion-network-for-skin-lesion-classification","From local textures to global attention - a hybrid feature fusion network for skin lesion classification","","Accurate detection of skin cancer from RGB images is difficult due to lesion appearance variability, differences in skin types, and limited clinical interpretability of models. A unified feature-fusion framework is proposed, combining normalization, class-aware selective augmentation, and feature-level fusion from VGG-16, ResNet-50 with dermatological enhancement, and a Vision Transformer using dynamic patching, followed by HDFNet classification. Experiments on ISIC 2019, ISIC 2020, PAD-UFES, and DermQuest DERMIS achieve 94.5% accuracy and 97% AUC-ROC, outperforming state-of-the-art approaches with GradCAM explanations.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/from-local-textures-to-global-attention-a-hybrid-feature-fusion-network-for-skin-lesion-classification/351834/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/from-local-textures-to-global-attention-a-hybrid-feature-fusion-network-for-skin-lesion-classification/351834.png","ImageObject",300,407,{"name":92,"@type":93},"Elsa","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-30","2026-09-22",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",12,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"Why is skin cancer detection from RGB images challenging?","Question",{"text":113,"@type":114},"Lesion appearance varies widely, skin types differ, and many models are hard to interpret clinically, which can cause false positives or false negatives.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"What does the proposed model combine to improve classification?",{"text":118,"@type":114},"It uses normalization and class-aware selective augmentation, then fuses features extracted from VGG-16, ResNet-50, and a Vision Transformer with dynamic patching, followed by HDFNet classification.",{"name":120,"@type":111,"acceptedAnswer":121},"Which datasets were used and what performance was achieved?",{"text":122,"@type":114},"The model was trained and tested on ISIC 2019, ISIC 2020, PAD-UFES, and DermQuest DERMIS, reaching 94.5% accuracy and 97% AUC-ROC.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},351834,1790334403,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":140,"language":141,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":142,"faqs":143,"seo_title":144,"seo_description":67,"update_tm":145,"read_time":146,"preview_page_count":19,"preview_urls":147},137455077381,"https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d","[www. nature.com/scientificreports](www. nature.com/scientificreports)  \nOPEN  \nFrom local textures to global attention: a hybrid feature fusion network for skin lesion classification  \nYazeedAlkhrijah1, Syed Nehal Hassan Shah2, Syed Muhammad Usman3,  \nShehzad Khalid4,3􀀍, Syed Abdullah Shah2, Sulieman S. Alshuhri5 & Jungpil Shin6  \nAccurate detection of skin cancer detection using RGB images remains a challenge due to  \nmultiple factors including variability in lesion appearance, difference in skin types, and the clinical interpretability of the models. To address these challenges, we present a unified feature-fusion framework that integrates deep learning methods for accurate classification of dermoscopic skin lesions. Our proposed method consists of preprocessing in which we performed normalization, class aware selective augmentation, followed by feature-level fusion from three customized deep learning architectures i.e.; VGG-16 with adaptive layer configuration, ResNet-50 with dermatological feature enhancement, and a Vision Transformer (ViT) with dynamic patching. These extracted features are then fused to form a comprehensive feature vector and classification is done using an HDFNet which is a two-layer Deep Neural Network (DNN). We have trained and tested the proposed model on four publicly available datasets including ISIC 2019, ISIC 2020, PAD-UFES, and DermQuest DERMIS. We achieved classification accuracy of 94.5% and anAUC-ROC of 97%. Our proposed method outperforms the existing state-of-the-art models and also provide interpretable predictions supported by GradCAM-based visual explanations.  \nThere is a significant increase in the number of Malignant melanoma and non-melanoma skin cancers (NMSC) . These two types of skin cancer have become one of the most important and rapidly expanding global public health issues worldwide1,2. Reducing skin cancer death rates requires early and accurate detection. The use of solarium, prolonged exposure to the sun, and excessive ultraviolet (UV) radiation are potential causes3,4. In this research, we have focused on four lesion categories that including Melanocytic Nevus (NV), Basal Cellular Carcinoma (BCC), Benign Keratosis-like Lesions (BKL) and Melanoma (MEL) . BCC and MEL are among the most prevalent types of skin cancer. Despite its prevalence, it develops gradually and rarely spreads to other parts of body. On the other hand, MEL is highly aggressive and invades different parts of the body, if not identified and treated on time1,2. Automated detection of skin cancer is a challenging task due to various recognizable patterns including skin types, noise, light intensity and angle etc in the skin cancer images that can result in either increased false positives or false negatives. Advancements in the Deep Learning (DL) methods5–20 have enabled automated detection of skin cancer with high accuracy14. Variety of deep learning architectures21, including Convolutional Neural Network (CNN), Deep Neural Network (DNN), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Residual Network (ResNet), Inception, Xception, and Visual Geometry Group-16 (VGG16), have been proposed by researchers in recent years for lesion detection and identification5,8, 12 but inceptionV3 and VGG16 have shown improved performance12. The recent developments of new models, that can extract low-level features without falling into any bias, have been useful in the domain of medical categorization of images1. Challenges of the visual inspection include skin type variations, presentation method of the affected region, and the inherent subjectivity that ensures a thorough dermatological analysis22. ResNet, VGG, Inception,  \n1Department of Electrical Engineering, Imam Mohammad ibn Saud Islamic University (IMSIU), 11623 Riyadh, Saudi Arabia. 2Department of Creative Technologies, Air University, Islamabad, Pakistan. 3Department of Computer Science, Bahria School of Engineering and Applied Sciences (BSEAS), Ba","cbCaifGgwCTZCpxL","https://ap.wps.com/l/cbCaifGgwCTZCpxL","pdf",4026956,22,"English","# Introduction\n## Challenges in skin cancer detection\n## Proposed hybrid feature fusion framework\n# Datasets and results\n## Public datasets used\n## Performance metrics and comparisons\n# Method overview\n## Preprocessing and augmentation\n## Feature extraction architectures","[{\"question\":\"Why is skin cancer detection from RGB images challenging?\",\"answer\":\"Lesion appearance varies widely, skin types differ, and many models are hard to interpret clinically, which can cause false positives or false negatives.\"},{\"question\":\"What does the proposed model combine to improve classification?\",\"answer\":\"It uses normalization and class-aware selective augmentation, then fuses features extracted from VGG-16, ResNet-50, and a Vision Transformer with dynamic patching, followed by HDFNet classification.\"},{\"question\":\"Which datasets were used and what performance was achieved?\",\"answer\":\"The model was trained and tested on ISIC 2019, ISIC 2020, PAD-UFES, and DermQuest DERMIS, reaching 94.5% accuracy and 97% AUC-ROC.\"}]","From local textures to global attention - a hybrid feature fusion network for skin lesion classification | PDF",1790096256,55,[148,152,156,160],{"page_no":8,"url":149,"width":150,"height":151},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/ea7ce1450456654952e711b276bb1da0.webp",983,1291,{"page_no":14,"url":153,"width":154,"height":155},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/278f4309f74ce0940fbb1bfa3d98ef6b.webp",982,1290,{"page_no":81,"url":157,"width":158,"height":159},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/282e6075ed36a8b30afb5b7e80564284.webp",981,1289,{"page_no":19,"url":161,"width":162,"height":163},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/b496c8cfa09de3989dd1f20e9095e84f.webp",980,1287]