[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-366598-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-366598-en":130},{"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":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","exploring-the-role-of-preprocessing-combinations-in-hyperspectral-imaging-for-deep-learning-colorectal-cancer-detection","Exploring the role of preprocessing combinations in hyperspectral imaging for deep learning colorectal cancer detection","","The study compares preprocessing techniques for hyperspectral deep learning diagnostics of colorectal cancer, focusing on scaling and noise reduction choices across spatial and spectral axes of hyperspectral datacubes, plus varying removal of blood and light reflections. It evaluates how patch size influences model performance and tests strategies to mitigate dataset imbalance where cancerous tissues are underrepresented. Results show standardization improves sensitivity and specificity over normalization, larger patches increase performance, while noise reduction can degrade it; blood filtering outperforms reflected-light filtering.",{"@graph":14,"@context":73},[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/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/exploring-the-role-of-preprocessing-combinations-in-hyperspectral-imaging-for-deep-learning-colorectal-cancer-detection/366598/",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/exploring-the-role-of-preprocessing-combinations-in-hyperspectral-imaging-for-deep-learning-colorectal-cancer-detection/366598.png","ImageObject",300,407,{"name":42,"@type":43},"Aldword","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-24","2026-09-23",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",9,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"Which preprocessing approach performed best for scaling hyperspectral data?","Question",{"text":63,"@type":64},"Standardization significantly improved both sensitivity and specificity compared with normalization.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"How did input patch size affect model performance?",{"text":68,"@type":64},"Larger input patch sizes enhanced performance by capturing more spatial context.",{"name":70,"@type":61,"acceptedAnswer":71},"What was the observed impact of noise reduction and reflection filtering?",{"text":72,"@type":64},"Noise reduction unexpectedly degraded performance; blood filtering was more effective than filtering reflected light pixels, and neither approach produced significant results under the tested conditions.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},366598,1790960886,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,106,111,115,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":25,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":107,"doc_module":4,"doc_module_name":25,"category_name":108,"show_sort_weight":109,"slug":110},7,"Healthcare",40,"healthcare",{"id":112,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":113,"slug":114},8,30,"research-report",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general",{"code":4,"msg":82,"data":131},{"doc_id":79,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":112,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":139,"language":140,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":12,"update_tm":144,"read_time":145,"preview_page_count":33,"preview_urls":146},2336478940917,"https://ap-avatar.wpscdn.com/davatar_6f874abed73319feea01a86fa6f0fab8","[www. nature.com/scientificreports](www. nature.com/scientificreports)  \nOPEN  \nExploring the role of preprocessing combinations in hyperspectral imaging for deep learning colorectal cancer detection  \nMariiaTkachenko1,2􀀍, Benjamin Huber3, Serhii Hamotskyi3, Boris Jansen-Winkeln4, Ines Gockel4,5, Thomas Neumuth2, Hannes Köhler2 & Marianne Maktabi2,3  \nThis study compares various preprocessing techniques for hyperspectral deep learning–based cancer diagnostics. The study considers different spectrum scaling and noise reduction options across spatial and spectral axes of hyperspectral datacubes, as well varying levels of blood and light reflections removal. We also examine how the size of the patches extracted from the hyperspectral data affects the models’ performance. We additionally explore various strategies to mitigate our dataset’s imbalance (where cancerous tissues are underrepresented) . Our results indicate that. Scaling:  \nStandardization significantly improves both sensitivity and specificity compared to Normalization. Larger input patch sizes enhance performance by capturing more spatial context. Noise reduction unexpectedly degrades performance. Blood filtering is more effective than filtering reflected light pixels, although neither approach produces significant results. By carefully maintaining consistent testing conditions, we ensure a fair comparison across preprocessing methods and reproducibility. Our findings highlight the necessity of careful preprocessing selection to maximize deep learning performance in medical imaging applications.  \nKeywords Preprocessing, Cancer classification, Colorectal cancer, Hyperspectral imaging, Machine learning, Convolutional networks  \nCancer is one of the leading causes of death worldwide1. Colorectal cancer —third by incidence and second by mortality2 — accounts for a significant portion of cancer-related morbidity. Techniques usable during preoperative3, intraoperative4, or postoperative stages (e.g., to detect cancer regions and ensure tumor-free resection margins5 can aid in cancer diagnosis and improve outcomes.  \nHyperspectral imaging (HSI) combines imaging and spectroscopy, creating image data containing both spatial information and the spectrum for each pixel, usually extending beyond the wavelengths seen by humans. HSI has shown potential in discriminating between tissue structures by analyzing their spectrum, thereby distinguishing healthy and pathological (e.g., cancerous) tissues in a contactless, radiation-free and non-invasive manner, and has seen increasing adoption in the medical field4.  \nHSI has been used for cancer detection in humans, for skin6,7, breast8, head and neck9, brain10, 11, oral12, gastric13 cancer, as well as colorectal cancer14, 15, and various others.  \nDifferent approaches exist for extracting this information from hypercubes: both conventional Machine Learning (ML) models, which have a stronger reliance on domain-specific knowledge to extract data, and Deep Learning (DL) ones, which can discover structures in high-dimensional data with very little engineering by hand16 and have obtained better classification performance than traditional approaches17. DL models used for HSI cancer detection have been primarily (but not exclusively) convolutional neural networks (CNNs)18–21, including 3D-CNNs14,22. Other approaches have been used with success as well, for instance semi-supervised Graph Convolutional Networks that leverage the use of unlabeled pixels23.  \n1Center for scalable data analytics and artificial intelligence (ScaDS.AI) Dresden/Leipzig, Leipzig, Germany.  \n2Innovation Center Computer-Assisted Surgery (ICCAS), University of Leipzig, Leipzig 04103, Germany. 3Anhalt University of Applied Sciences, Köthen 06366, Germany. 4Department of Visceral, Transplant, Thoracic and Vascular Surgery, University Hospital of Leipzig, Leipzig 04103, Germany. 5Department of Gastrointestinal Surgery, IRCCS San Raffaele Scientific Institute and San Raffaele","cbCaiggyrkRfAmvk","https://ap.wps.com/l/cbCaiggyrkRfAmvk","pdf",4848191,15,"English","# Study objective and problem context\n## Hyperspectral imaging and deep learning for cancer detection\n# Experimental design and preprocessing factors\n## Scaling, noise reduction, and reflections removal\n## Patch size and dataset imbalance handling\n# Results and implications","[{\"question\":\"Which preprocessing approach performed best for scaling hyperspectral data?\",\"answer\":\"Standardization significantly improved both sensitivity and specificity compared with normalization.\"},{\"question\":\"How did input patch size affect model performance?\",\"answer\":\"Larger input patch sizes enhanced performance by capturing more spatial context.\"},{\"question\":\"What was the observed impact of noise reduction and reflection filtering?\",\"answer\":\"Noise reduction unexpectedly degraded performance; blood filtering was more effective than filtering reflected light pixels, and neither approach produced significant results under the tested conditions.\"}]","Exploring the role of preprocessing combinations in hyperspectral imaging for deep learning colorectal cancer detection | PDF",1790166358,38,[147,151,155,159],{"page_no":22,"url":148,"width":149,"height":150},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/e7ad12f4babfe66788f9a02e68411d26.webp",983,1291,{"page_no":26,"url":152,"width":153,"height":154},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/9c946131cd35fd5487d3d6852810dea5.webp",982,1290,{"page_no":30,"url":156,"width":157,"height":158},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/57d46bf54213df263e0bdec5a62b8801.webp",981,1289,{"page_no":33,"url":160,"width":161,"height":162},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-16/1b927873507113494e383d803e29c60f.webp",980,1287]