[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-367488-105":59,"doc-detail-367488-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","screening-of-oral-potentially-malignant-disorders-and-oral-cancer-using-deep-learning-models","Screening of oral potentially malignant disorders and oral cancer using deep learning models","","Oral cancer, preventable yet often diagnosed late, carries high mortality and reduces quality of life. This study evaluates AI screening tools for oral potentially malignant disorders (OPMDs) and oral cancer via native or cloud web applications on smartphone devices. Two deep learning models, DenseNet201 and an adapted FixCaps based on capsule networks, were trained and tested on 518 oral-cavity images with standardized lesion annotation and classification.",{"@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/screening-of-oral-potentially-malignant-disorders-and-oral-cancer-using-deep-learning-models/367488/",{"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/screening-of-oral-potentially-malignant-disorders-and-oral-cancer-using-deep-learning-models/367488.png","ImageObject",300,407,{"name":92,"@type":93},"Taylor Morgan","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-29","2026-09-23",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},"What problem does this work address in oral cancer screening?","Question",{"text":113,"@type":114},"Late detection of oral cancer leads to high mortality and poorer quality of life. The study focuses on enabling earlier diagnosis through AI screening for OPMDs and oral cancers.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"Which deep learning models were evaluated and how were they trained?",{"text":118,"@type":114},"DenseNet201 and an adapted FixCaps model were trained and tested on 518 oral-cavity images. DenseNet201 used a pre-trained model, while FixCaps was modified from capsule networks and trained from the ground up.",{"name":120,"@type":111,"acceptedAnswer":121},"How did the two models perform in lesion classification?",{"text":122,"@type":114},"DenseNet201 achieved an F1 score of 87.50% and AUC of 0.97, while FixCaps achieved an F1 score of 82.8% and AUC of 0.93.","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},367488,1790208358,{"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":44,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145,"preview_page_count":19,"preview_urls":146},1099523885336,"https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c","[www. nature.com/scientificreports](www. nature.com/scientificreports)  \nOPEN  \nScreening of oral potentially malignant disorders and oral cancer using deep learning models  \nKarishma Madhusudan Desai1,2, Pragya Singh2􀀍, Mahima Smriti2, VivekTalwar3, Manav Chaudhary2, George Paul2, Subhas Chandra Kolli2, Parisa Sai Raghava2, GollaVamshi Krishna2, C. V. Jawahar3, P. K. Vinod4,5, Varma Konala2,6 & Ramanathan Sethuraman6  \nOral cancer though preventable, shows high mortality and affect the overall quality of life when detected in late stages. Screening techniques that enable early diagnosis are the need of the hour. The present work aims to evaluate the effectiveness of AI screening tools in the diagnosis of OPMDs and Oral cancers via native or web-application (cloud) using smart phone devices. We trained and tested two deep learning models namely DenseNet201 and FixCaps using 518 images of the oral cavity. While DenseNet201 is a pre-trained model, we modified the FixCaps model from capsule network and trainedit ground up. Standardized protocols were used to annotate and classify the lesions (suspicious vs. nonsuspicious) . In terms of model performance, DenseNet201 achieved an F1 score of 87.50% andAUC of 0.97; while FixCaps exhibited F1 score of 82.8% andAUC of 0.93. DenseNet201 model (20 M) serves as a robust screening model (accuracy 88.6%) that can be hosted on a web-application in the cloud servers; while the adapted FixCaps model with its low parameter size of 0.83 M exhibits comparable accuracy (83.8%) allowing easy transitioning into a native phone-based screening application.  \nKeywords Screening, Oral cancer, Oral potentially malignant disorders, Deep learning, Web application  \nOral and lip cancers, though preventable, are associated with 1,77,757 deaths annually1,2. The low- and middle-income countries have witnessed snowballing incidence of oral cancers and their precursors lesions‘oral potentially malignant disorders (OPMDs)’. Screening via clinical oral examination (COE) is considered as the most effective method for early detection of OPMDs and oral cancer1. Additionally, as visual screening is cost-effective, researchers and governing bodies globally have laid emphasis to promote and implement conventional oral screening programs3–5. Though early detection of OPMDs and early-stage cancers can reduce patient mortality6, there are several roadblocks such as heterogeneity in examination techniques, diverse clinical presentations and lack of health literacy. Apart from the known grave health consequences and morbidity, treatment of advanced carcinomas adds enormous economic and psychological burden on the affected individual7.  \nIn several developing countries, besides screening and diagnosis, the access to healthcare services remains a pertinent challenge. This is further aggravated by urban–rural divide in low-and middle-income countries (LMIC) like India, wherein two-thirds of the population reside in rural areas which have limited accessibility and affordability to health services8. Recently, community health workers in India such as Accredited social health activists (ASHA) have helped to address this challenge. ASHA’s have been trained to perform screening of the oral cavity and capture images of oral cavity using smartphones. These images when uploaded to AI applications can guide the health workers to refer high-risk patients to specialists for timely management9–11. AI models have shown value in early detection of OPMDs and oral cancers11–13.  \nAdvanced computer vision algorithms based on deep learning help recognize, process and present desirable output based on the quality of images, annotations and data reconstruction techniques. Clinicians globally have put together one such effort via the Mobile Mouth Screening Anywhere (MeMoSA) project and deep learning based convolutional neural networks (CNN)14. They employed the box annotation approach with residual  \n1Tokyo Dental College, 2 Chome-9-18 Misakicho, ","cbCaigdXbN8LqegO","https://ap.wps.com/l/cbCaigdXbN8LqegO","pdf",8013014,"English","# Introduction\n## Clinical need and screening challenges\n## AI-enabled remote screening in LMICs\n## Related deep learning approaches\n# Methods and Models\n## Data annotation and lesion classification\n## DenseNet201 and FixCaps model design\n# Results\n## Performance metrics (F1 score and AUC)\n# Discussion\n## Deployment constraints and adoption barriers","[{\"question\":\"What problem does this work address in oral cancer screening?\",\"answer\":\"Late detection of oral cancer leads to high mortality and poorer quality of life. The study focuses on enabling earlier diagnosis through AI screening for OPMDs and oral cancers.\"},{\"question\":\"Which deep learning models were evaluated and how were they trained?\",\"answer\":\"DenseNet201 and an adapted FixCaps model were trained and tested on 518 oral-cavity images. DenseNet201 used a pre-trained model, while FixCaps was modified from capsule networks and trained from the ground up.\"},{\"question\":\"How did the two models perform in lesion classification?\",\"answer\":\"DenseNet201 achieved an F1 score of 87.50% and AUC of 0.97, while FixCaps achieved an F1 score of 82.8% and AUC of 0.93.\"}]","Screening of oral potentially malignant disorders and oral cancer using deep learning models | PDF",1790171032,23,[147,151,155,159],{"page_no":8,"url":148,"width":149,"height":150},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/883bc9a5ec7f6e9bb843abc68e51e788.webp",983,1291,{"page_no":14,"url":152,"width":153,"height":154},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/beaec8c6146f4eeddee655864f0fc23a.webp",982,1290,{"page_no":81,"url":156,"width":157,"height":158},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/4bc809636aa99a96ddc68b19a7e3a61b.webp",981,1289,{"page_no":19,"url":160,"width":161,"height":162},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/b349d773bb1e223f91e76ba1e0de75bd.webp",980,1287]