[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-291248-105":59,"doc-detail-291248-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","handct-hands-on-computational-dataset-for-x-ray-computed-tomography-and-machine-learning","HandCT - hands-on computational dataset for X-Ray Computed Tomography and Machine-Learning","","Machine-learning methods depend on sufficiently large datasets to learn data distributions, yet X-Ray Computed Tomography (XCT) often lacks scalable resources for realistic 3D problems. HandCT provides a 3D procedural, open-source dataset built from a left-hand meshed model plus Python scripts that randomize anatomy and pose while preserving realistic features. The workflow uses Blender and Blender’s Python core for pose and mesh generation, while Python converts files to matrices and prepares volumes. The lightweight dataset enables truthful samples, version-controlled enhancements, and motivates future anatomically accurate extensions.",{"@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/handct-hands-on-computational-dataset-for-x-ray-computed-tomography-and-machine-learning/291248/",{"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/handct-hands-on-computational-dataset-for-x-ray-computed-tomography-and-machine-learning/291248.png","ImageObject",300,407,{"name":92,"@type":93},"Ophelia","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-26","2026-09-17",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",16,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"Why are existing X-Ray CT datasets insufficient for machine learning on realistic 3D problems?","Question",{"text":113,"@type":114},"Existing datasets are limited by their design purposes and the difficulty of collecting and using CT data, which prevents scalable training for realistic 3D tasks.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"What is HandCT and how are its samples generated?",{"text":118,"@type":114},"HandCT is a 3D procedural dataset that starts from a meshed model of a left hand and uses scripts to randomly change anatomical properties and pose while conserving realistic features.",{"name":120,"@type":111,"acceptedAnswer":121},"Which tools does HandCT rely on for dataset generation and processing?",{"text":122,"@type":114},"It is an open-source solution built on Blender for modeling and pose generation, and Python for file conversion to matrices and for scaling/centering volumes for further processing.","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},291248,1789643867,{"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":34,"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":130,"read_time":144,"preview_page_count":4},7971461741311,"https://ap-avatar.wpscdn.com/avatar/74000253aff267980c6?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779345379180704826","arXiv :2304 . 14412v1 [ ee ss .IV] 17 Apr 2023  \nHandCT: hands-on computational dataset for X-Ray Computed Tomography and Machine-Learning  \nValat Emilien, Valat Loth  \nMay 1, 2023  \nAbstract  \nMachine-learning methods rely on su􀀎ciently large dataset to learn data distributions. They are widely used in research in X-Ray Computed Tomography, from low-dose scan denoising to optimisation of thereconstruction process. The lack of datasets prevents the scalability of these methods to realistic 3D problems. We develop a 3D procedural dataset in order to produce samples for data-driven algorithms. It is made of a meshed model of a left hand and a script to randomly change its anatomic properties and pose whilst conserving realistic features. This open-source solution relies on the freeware Blender and its Python core. Blender handles the modelling, the mesh and the generation of the hand's pose, whilst Python processes 􀀌le format conversion from obj 􀀌le to matrix and functions to scale and center the volume for further processing. Dataset availability and quality drives research in machine-learning. We design a dataset that weighs few megabytes, provides truthful samples and proposes continuous enhancements using version control. We anticipate this work to be a starting point for anatomically accurate procedural datasets. For instance, by adding more internal features and 􀀌ne tuning their X-Ray attenuation properties.  \n1 Introduction  \nBackground & Summary  \nMachine-learning (ML) methods are widely used in X-Ray Computed Tomography (XCT), mainly for low-dose denoising, scarce-view CT and dual-energy CT. They rely on data samples to be trained on, and their performance is conditioned by the availability and quality of the datasets. At present, the limitations to the existing datasets we identify are related to the purposes for which they were designed as well as the di􀀎culty to collect and use CT data.  \nFigure 1: Example of data-leakage for 2D slices of a 3D volume. The color indicate the level of correlation with the last few training samples. Green, orange and red indicate strong, weak, poor correlation, respectively.  \nBenchmarking reconstruction algorithms and training ML approaches require di􀀋erent types of datasets. The Shepp-Logan phantom [14] is a computational phantom used routinely when comparing reconstruction processes. It consists of a set of ellipses with various X-Ray attenuation values and radius, and has been designed to resemble to a human brain. As training a ML algorithm on one data point is irrelevant, it cannot be used for that purpose. [6] consists in a set of projections and an associated reconstruction of a single walnut, chosen for its likeness to a human brain. As all these items relate to the same data-point, it is unsuitable for ML purposes as well. The SophiaBeads dataset [2] is used to benchmark reconstruction processes against a certain level of angular scarcity in the projection set. Even though there are di􀀋erent data-points in the angular domain, all the measures relate to one actual sample. [16, 3, 15] have the same limitation, knowing that the number of samples is to little to undertake ML procedures without over-􀀌tting the data. One might consider using those datasets for 2D CT by reconstructing the 3D volumes, slice and re-project them to produce as many samples as there are slices. This can lead to data-leakage and should be avoided, as illustrated in Fig. 1.  \nWhen training image improvement methods on the 􀀌rst 180 slices out of 200 that compose a volume from the Sophiabeads dataset, the samples are highly correlated with their immediate neighbours. As such, when evaluating on the last 10% of the dataset which corresponds to the last 20 slices, we observe that the performance of an approach depends on the index of the slice considered: the closer to the last slice of the training set a test sample is, the better. Averaging the performance across the test dataset might mask this, but is not v","cbCaigvWXMiPOsCM","https://ap.wps.com/l/cbCaigvWXMiPOsCM","pdf",2882844,"English","# Introduction\n## Background & Summary","[{\"question\":\"Why are existing X-Ray CT datasets insufficient for machine learning on realistic 3D problems?\",\"answer\":\"Existing datasets are limited by their design purposes and the difficulty of collecting and using CT data, which prevents scalable training for realistic 3D tasks.\"},{\"question\":\"What is HandCT and how are its samples generated?\",\"answer\":\"HandCT is a 3D procedural dataset that starts from a meshed model of a left hand and uses scripts to randomly change anatomical properties and pose while conserving realistic features.\"},{\"question\":\"Which tools does HandCT rely on for dataset generation and processing?\",\"answer\":\"It is an open-source solution built on Blender for modeling and pose generation, and Python for file conversion to matrices and for scaling/centering volumes for further processing.\"}]","HandCT - hands-on computational dataset for X-Ray Computed Tomography and Machine-Learning | PDF",18]