{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T22:10:45Z","timestamp":1780611045880,"version":"3.54.1"},"reference-count":46,"publisher":"Wiley","issue":"7","license":[{"start":{"date-parts":[[2025,7,15]],"date-time":"2025-07-15T00:00:00Z","timestamp":1752537600000},"content-version":"vor","delay-in-days":14,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"}],"content-domain":{"domain":["aapm.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Medical Physics"],"published-print":{"date-parts":[[2025,7]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>The judicious use of CT in pediatric cardiac applications is warranted because young patients face the need for repeated imaging and increased lifetime cancer risk after ionizing radiation exposure. The quality of pediatric cardiac CT scans is variable because of limited protocols optimizations for pediatric patients, the common presence of metallic implants following treatment, and disparities in denoising algorithm performance between adult and pediatric scans. Two recent technological developments promise to improve the average quality of pediatric CT scans at fixed or reduced dose: clinical photon\u2010counting CT (PCCT) and deep learning (DL) algorithms for CT image denoising. Given advancements to accommodate variable image quality, these technologies will deliver improved spatial resolution, noise performance, and contrast resolution for pediatric cardiac CT imaging.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>To advance self\u2010supervised DL denoising methods to accommodate variable image quality in pediatric cardiac CT data.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Starting with the popular Vision Transformer (ViT) DL architecture, two targeted architectural changes were made: (1) the multi\u2010layer perceptrons (MLPs) were modified to allow cross\u2010token recombination of encoded image data following attention computations (parallels patch\u2010wise weighting and averaging in non\u2010local means [NLM]), and (2) the network head was replaced with the equivalent of an overcomplete dictionary to perform dictionary sparse coding (SC). This modified, 3D ViT (mViT) was then trained in a dynamic fashion: the balance between data fidelity and representation sparsity was adjusted during training such that the average fidelity error remained consistent with localized estimates of image noise. To demonstrate the newly proposed method, the mViT was trained with pediatric cardiac photon\u2010counting x\u2010ray CT data with variable levels of image noise (NAEOTOM Alpha PCCT scanner; retrospective data from 20 patients scanned at Duke University; ages: 1\u201318 years; iterative reconstruction noise level in the left ventricle: 20\u201355 HU). Data from one patient with the highest levels of noise was reserved for validation. Testing data included Alpha data from three additional Duke patients (2\u00a0&lt;\u00a01 year old) and a murine cardiac PCCT data set acquired on a preclinical system.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The validation denoising results demonstrate that SC with the mViT preserves anatomic structures relevant to the diagnosis and treatment of congenital heart defects (coronary artery origins; valve leaflets; left ventricle boundaries) while achieving similar intensity bias and lower intensity variance values than competing denoising methods (bilateral filtration [BF], NLM, dictionary SC, block matching 4D, orthogonal matching pursuit, Noise2Void). Applying the trained mViT network to preclinical PCCT demonstrated robust generalization performance to high levels of image noise (\u223c230 HU) and differing image contrast; however, applying the network to clinical PCCT data in younger patients (&lt;\u00a01 year old) demonstrated some smoothing of image details in data already heavily denoised during reconstruction.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>This work demonstrates robust, self\u2010supervised denoising of pediatric cardiac PCCT data through data adaptation during network training based on local noise estimates. The trained network generalizes to data sets with high levels of noise and differing image contrast relative to the training data, suggesting that self\u2010supervised fine tuning may allow the trained network to address related CT denoising problems.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1002\/mp.17918","type":"journal-article","created":{"date-parts":[[2025,7,15]],"date-time":"2025-07-15T07:59:01Z","timestamp":1752566341000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":2,"title":["Denoising pediatric cardiac photon\u2010counting CT data with sparse coding and data\u2010adaptive, self\u2010supervised deep learning"],"prefix":"10.1002","volume":"52","author":[{"given":"Darin P.","family":"Clark","sequence":"first","affiliation":[{"name":"Quantitative Imaging and Analysis Lab Department of Radiology Duke University  Durham North Carolina USA"},{"name":"Center for Virtual Imaging Trials Carl E. Ravin Advanced Imaging Laboratories Department of Radiology Duke University  Durham North Carolina USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joseph Y.","family":"Cao","sequence":"additional","affiliation":[{"name":"Department of Radiology Division of Pediatric Radiology Duke University  Durham North Carolina USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cristian T.","family":"Badea","sequence":"additional","affiliation":[{"name":"Quantitative Imaging and Analysis Lab Department of Radiology Duke University  Durham North Carolina USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2025,7,15]]},"reference":[{"key":"e_1_2_7_2_1","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.2018172656"},{"key":"e_1_2_7_3_1","doi-asserted-by":"publisher","DOI":"10.2214\/AJR.22.28391"},{"key":"e_1_2_7_4_1","doi-asserted-by":"publisher","DOI":"10.1007\/s10554-020-02118-7"},{"key":"e_1_2_7_5_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.16901"},{"key":"e_1_2_7_6_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2024.108112"},{"key":"e_1_2_7_7_1","doi-asserted-by":"publisher","DOI":"10.1109\/TRPMS.2023.3341903"},{"issue":"110","key":"e_1_2_7_8_1","first-page":"3371","article-title":"Stacked denoising autoencoders: learning useful representations in a deep network with a local denoising criterion","volume":"11","author":"Vincent P","year":"2010","journal-title":"J Mach Learn Res"},{"key":"e_1_2_7_9_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2018.01.015"},{"key":"e_1_2_7_10_1","doi-asserted-by":"crossref","unstructured":"ClarkD SchwartzF CaoJ BadeaCT.Denoising pediatric cardiac photon\u2010counting CT data using volumetric vision transformers and unpaired training data. In:Medical Imaging 2024: Physics of Medical Imaging. Vol12925.Proceedings of SPIE;2024:12925\u201012.","DOI":"10.1117\/12.3006863"},{"key":"e_1_2_7_11_1","doi-asserted-by":"crossref","unstructured":"MoranN SchmidtD ZhongY CoadyP.Noisier2noise: learning to denoise from unpaired noisy data. In:Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition;2020:12064\u201012072.","DOI":"10.1109\/CVPR42600.2020.01208"},{"key":"e_1_2_7_12_1","first-page":"864","article-title":"Noise2score: tweedie's approach to self\u2010supervised image denoising without clean images","volume":"34","author":"Kim K","year":"2021","journal-title":"Adv Neural Inf Process Syst"},{"key":"e_1_2_7_13_1","unstructured":"LehtinenJ MunkbergJ HasselgrenJ et\u00a0al.Noise2Noise: Learning image restoration without clean data.arXiv. Published March 12 2018. arXiv:1803.04189."},{"key":"e_1_2_7_14_1","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/ac8866"},{"key":"e_1_2_7_15_1","doi-asserted-by":"crossref","unstructured":"KrullA BuchholzT\u2010O JugF.Noise2void\u2010learning denoising from single noisy images. In:Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition;2019:2124\u20102132.","DOI":"10.1109\/CVPR.2019.00223"},{"key":"e_1_2_7_16_1","doi-asserted-by":"crossref","unstructured":"H\u00f6ckE BuchholzT\u2010O BrachmannA JugF FreytagA.N2v2\u2010fixing noise2void checkerboard artifacts with modified sampling strategies and a tweaked network architecture. In:European Conference on Computer Vision;2022:722\u2010738.","DOI":"10.1007\/978-3-031-25069-9_33"},{"key":"e_1_2_7_17_1","doi-asserted-by":"publisher","DOI":"10.1109\/TCI.2021.3050266"},{"key":"e_1_2_7_18_1","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2019.2934178"},{"key":"e_1_2_7_19_1","doi-asserted-by":"crossref","unstructured":"GondaraL Medical image denoising using convolutional denoising autoencoders.Paper presented at: 2016 IEEE 16th international conference on data mining workshops (ICDMW) 2016.","DOI":"10.1109\/ICDMW.2016.0041"},{"issue":"1","key":"e_1_2_7_20_1","first-page":"2887","article-title":"Convolutional neural networks analyzed via convolutional sparse coding","volume":"18","author":"Papyan V","year":"2017","journal-title":"J Mach Learn Res"},{"key":"e_1_2_7_21_1","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.212579"},{"key":"e_1_2_7_22_1","article-title":"An image is worth 16\u00d716 words: transformers for image recognition at scale","author":"Dosovitskiy A","year":"2020","journal-title":"arXiv"},{"key":"e_1_2_7_23_1","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-007-0052-1"},{"key":"e_1_2_7_24_1","doi-asserted-by":"crossref","unstructured":"WangX GirshickR GuptaA HeK.Non\u2010local neural networks. In:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition;2018:7794\u20107803.","DOI":"10.1109\/CVPR.2018.00813"},{"key":"e_1_2_7_25_1","doi-asserted-by":"publisher","DOI":"10.1109\/TSP.2006.881199"},{"key":"e_1_2_7_26_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0218417"},{"key":"e_1_2_7_27_1","doi-asserted-by":"crossref","unstructured":"BadeaC ClarkD AlphinA et\u00a0al.Co\u2010clinical photon counting CT research for multi\u2010contrast imaging. In:Proceedings of the 7th International Conference on Image Formation in X\u2010Ray Computed Tomography;2022.","DOI":"10.1117\/12.2646399"},{"key":"e_1_2_7_28_1","doi-asserted-by":"publisher","DOI":"10.3390\/tomography9040102"},{"key":"e_1_2_7_29_1","doi-asserted-by":"publisher","DOI":"10.1137\/S003614450037906X"},{"key":"e_1_2_7_30_1","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.1995.10476626"},{"key":"e_1_2_7_31_1","doi-asserted-by":"crossref","unstructured":"ClarkD SchwartzF EulerA et\u00a0al.Unsupervised learning of robust models for cardiac and photon\u2010counting x\u2010ray CT denoising. In:Medical Imaging 2023: Physics of Medical Imaging. Vol13640.Proceedings of SPIE;2023:13640\u201021.","DOI":"10.1117\/12.2653640"},{"key":"e_1_2_7_32_1","article-title":"A method for stochastic optimization","author":"Kingma DP","year":"2014","journal-title":"arXiv"},{"key":"e_1_2_7_33_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.16532"},{"key":"e_1_2_7_34_1","doi-asserted-by":"crossref","unstructured":"TomasiC ManduchiR.Bilateral filtering for gray and color images. In:Proceedings of the Sixth International Conference on Computer Vision;1998:839\u2010846.","DOI":"10.1109\/ICCV.1998.710815"},{"key":"e_1_2_7_35_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2012.2210725"},{"key":"e_1_2_7_36_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.12174"},{"key":"e_1_2_7_37_1","doi-asserted-by":"publisher","DOI":"10.1007\/s00330-023-09644-7"},{"issue":"8","key":"e_1_2_7_38_1","first-page":"1","article-title":"Efficient implementation of the K\u2010SVD algorithm using batch orthogonal matching pursuit","volume":"40","author":"Rubinstein R","year":"2008","journal-title":"Cs Technion"},{"key":"e_1_2_7_39_1","doi-asserted-by":"publisher","DOI":"10.1161\/ATVBAHA.112.300717"},{"key":"e_1_2_7_40_1","doi-asserted-by":"publisher","DOI":"10.1148\/rg.295095002"},{"key":"e_1_2_7_41_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0235751"},{"key":"e_1_2_7_42_1","doi-asserted-by":"publisher","DOI":"10.1093\/ejcts\/ezx208"},{"key":"e_1_2_7_43_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0180324"},{"key":"e_1_2_7_44_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2014.2380993"},{"key":"e_1_2_7_45_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.15952"},{"key":"e_1_2_7_46_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2018.2827462"},{"key":"e_1_2_7_47_1","doi-asserted-by":"crossref","unstructured":"LiuZ LinY CaoY et\u00a0al.Swin transformer: hierarchical vision transformer using shifted windows. In:Proceedings of the IEEE\/CVF International Conference on Computer Vision;2021:10012\u201010022.","DOI":"10.1109\/ICCV48922.2021.00986"}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.17918","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,8]],"date-time":"2025-12-08T10:33:02Z","timestamp":1765189982000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.17918"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7]]},"references-count":46,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2025,7]]}},"alternative-id":["10.1002\/mp.17918"],"URL":"https:\/\/doi.org\/10.1002\/mp.17918","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,7]]},"assertion":[{"value":"2024-10-08","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-05-15","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-07-15","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e17918"}}