{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T14:18:57Z","timestamp":1788272337797,"version":"build-2803163510"},"reference-count":52,"publisher":"Wiley","issue":"8","license":[{"start":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T00:00:00Z","timestamp":1785715200000},"content-version":"vor","delay-in-days":2,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T00:00:00Z","timestamp":1785542400000},"content-version":"tdm","delay-in-days":0,"URL":"http:\/\/doi.wiley.com\/10.1002\/tdm_license_1.1"}],"funder":[{"DOI":"10.13039\/100019266","name":"Korea Medical Device Development Fund","doi-asserted-by":"publisher","award":["RS\u20102020\u2010KD000001"],"award-info":[{"award-number":["RS\u20102020\u2010KD000001"]}],"id":[{"id":"10.13039\/100019266","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100003725","name":"National Research Foundation of Korea","doi-asserted-by":"publisher","award":["RS\u20102025\u201000513383"],"award-info":[{"award-number":["RS\u20102025\u201000513383"]}],"id":[{"id":"10.13039\/501100003725","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["aapm.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Medical Physics"],"published-print":{"date-parts":[[2026,8]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Speckle patterns in ultrasound images often obscure anatomical details, leading to diagnostic uncertainty. Recently, various deep learning\u2010based techniques have been introduced to effectively suppress speckle; however, their high computational costs pose challenges for low\u2010resource devices, such as portable ultrasound systems.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>To address this issue, we introduce Edge Speckle Reduction and Image Enhancement (EdgeSRIE), a lightweight hybrid deep learning framework for real\u2010time speckle reduction and image enhancement in portable ultrasound imaging.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>\n                      The proposed framework consists of an unsupervised despeckling branch and a self\u2010supervised deblurring branch trained separately in sequence with AdamW (learning rate 1 \u00d7 10\n                      <jats:sup>\u2212<\/jats:sup>\n                      <jats:sup>4<\/jats:sup>\n                      ) and an\n                      <jats:italic>L2<\/jats:italic>\n                      loss. The integrated model was converted to an 8\u2010bit deployment model using post\u2010training quantization on a low\u2010resource system\u2010on\u2010chip. Training used 1779 B\u2010mode images (BUSI and HC18), validation used 77 raw acquisitions (PICMUS and CUBDL), and external evaluation used 94 EdgeFlow UH\u201010 bladder images. EdgeSRIE was compared with OSRAD, OBNLM, DIAE, DUNet, BRUNet, and USNet using contrast\u2010to\u2010noise ratio (CNR), speckle signal\u2010to\u2010noise ratio (SSNR), average gradient magnitude (AGM), and structural similarity index measure (SSIM); paired inference used two\u2010sided Wilcoxon signed\u2010rank tests with Holm correction and rank\u2010biserial and Hedges\u2019\n                      <jats:italic>g<\/jats:italic>\n                      effect sizes.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      Across the four representative cases, post\u2010training\u2010quantized EdgeSRIE achieved the largest mean improvement in CNR (64.9\u00a0\u00b1\u00a021.0%) and the smallest mean decrease in AGM (\u221216.6\u00a0\u00b1\u00a033.2%). SSNR (109.8\u00a0\u00b1\u00a045.8%) was in the OBNLM\u2010led range, whereas SSIM remained slightly lower than OSRAD and OBNLM on its native [0, 1] scale. The deployed model contained 17.67K parameters and reached 64.10 frames\/s on the target hardware. In the case\u2010matched analysis (\n                      <jats:italic>n<\/jats:italic>\n                      \u00a0=\u00a08), EdgeSRIE showed Holm\u2010corrected significant advantages over all six baselines in CNR, over five of six in SSNR, and over four of six in AGM. For SSIM, EdgeSRIE was significantly higher than the four deep learning baselines but lower than OSRAD and OBNLM. Across the statistically significant comparisons, the magnitude\u2010based effect sizes were large (Hedges\u2019\n                      <jats:italic>g<\/jats:italic>\n                      \u2265 0.8).\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>These results support the feasibility of EdgeSRIE as a compact, deployment\u2010oriented framework that balances speckle suppression, structural preservation, and real\u2010time execution for portable ultrasound imaging.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1002\/mp.70620","type":"journal-article","created":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T12:32:47Z","timestamp":1785760367000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["A hybrid deep learning framework for real\u2010time speckle reduction and image enhancement on portable ultrasound systems"],"prefix":"10.1002","volume":"53","author":[{"given":"Hyunwoo","family":"Cho","sequence":"first","affiliation":[{"name":"Department of Electronic Engineering Sogang University  Seoul South Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jaeseok","family":"Lee","sequence":"additional","affiliation":[{"name":"Department of Electronic Engineering Sogang University  Seoul South Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jongsoo","family":"Lee","sequence":"additional","affiliation":[{"name":"Department of Urology, Urological Science Institute, College of Medicine Yonsei University  Seoul South Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yangmo","family":"Yoo","sequence":"additional","affiliation":[{"name":"Department of Electronic Engineering Sogang University  Seoul South Korea"},{"name":"Department of Biomedical Engineering Sogang University  Seoul South Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jinbum","family":"Kang","sequence":"additional","affiliation":[{"name":"Department of Biomedical Software Engineering The Catholic University of Korea  Bucheon South Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2026,8,3]]},"reference":[{"key":"e_1_2_8_2_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2006.877092"},{"key":"e_1_2_8_3_1","doi-asserted-by":"publisher","DOI":"10.1364\/JOSA.66.001145"},{"key":"e_1_2_8_4_1","doi-asserted-by":"publisher","DOI":"10.1097\/00013644\u2010199300000\u201000001"},{"key":"e_1_2_8_5_1","doi-asserted-by":"publisher","DOI":"10.1177\/016173468801000201"},{"key":"e_1_2_8_6_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.irbm.2014.05.003"},{"key":"e_1_2_8_7_1","doi-asserted-by":"publisher","DOI":"10.1109\/58.285465"},{"key":"e_1_2_8_8_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jacc.2005.02.082"},{"key":"e_1_2_8_9_1","doi-asserted-by":"publisher","DOI":"10.1109\/T\u2010SU.1978.30978"},{"key":"e_1_2_8_10_1","doi-asserted-by":"publisher","DOI":"10.1109\/T\u2010SU.1983.31404"},{"key":"e_1_2_8_11_1","doi-asserted-by":"publisher","DOI":"10.1364\/AO.21.001157"},{"key":"e_1_2_8_12_1","doi-asserted-by":"publisher","DOI":"10.1007\/s13534\u2010014\u20100122\u20106"},{"key":"e_1_2_8_13_1","doi-asserted-by":"publisher","DOI":"10.1016\/S0146\u2010664X(81)80005\u20106"},{"key":"e_1_2_8_14_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.1982.4767223"},{"key":"e_1_2_8_15_1","doi-asserted-by":"publisher","DOI":"10.1109\/TASSP.1987.1165131"},{"key":"e_1_2_8_16_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2002.804276"},{"key":"e_1_2_8_17_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2007.891803"},{"key":"e_1_2_8_18_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2014.2371244"},{"key":"e_1_2_8_19_1","doi-asserted-by":"publisher","DOI":"10.1109\/TUFFC.2005.1561621"},{"key":"e_1_2_8_20_1","doi-asserted-by":"publisher","DOI":"10.1137\/040616024"},{"key":"e_1_2_8_21_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2009.2024064"},{"key":"e_1_2_8_22_1","doi-asserted-by":"publisher","DOI":"10.1109\/LSP.2018.2858147"},{"key":"e_1_2_8_23_1","doi-asserted-by":"crossref","unstructured":"HeK ZhangX RenS SunJ Deep residual learning for image recognition. In:Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).2016:770\u2010778.","DOI":"10.1109\/CVPR.2016.90"},{"key":"e_1_2_8_24_1","unstructured":"GoodfellowI Pouget\u2010AbadieJ MirzaM et\u00a0al.Generative adversarial nets. Advances in Neural Information Processing Systems.2014:27."},{"key":"e_1_2_8_25_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0205390"},{"key":"e_1_2_8_26_1","doi-asserted-by":"publisher","DOI":"10.1109\/TUFFC.2019.2903795"},{"key":"e_1_2_8_27_1","article-title":"Removal of speckle noises from ultrasound images using five different deep learning networks","volume":"29","author":"Karaoglu O","year":"2022","journal-title":"Eng Sci Technol Int J"},{"key":"e_1_2_8_28_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2022.106385"},{"key":"e_1_2_8_29_1","doi-asserted-by":"publisher","DOI":"10.14366\/usg.24005"},{"key":"e_1_2_8_30_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ultras.2024.107408"},{"key":"e_1_2_8_31_1","doi-asserted-by":"publisher","DOI":"10.1145\/3578938"},{"key":"e_1_2_8_32_1","doi-asserted-by":"publisher","DOI":"10.1055\/a-0783-2303"},{"key":"e_1_2_8_33_1","doi-asserted-by":"publisher","DOI":"10.1109\/TBME.2019.2912986"},{"key":"e_1_2_8_34_1","doi-asserted-by":"publisher","DOI":"10.24908\/pocus.v7i1.15052"},{"key":"e_1_2_8_35_1","doi-asserted-by":"publisher","DOI":"10.3390\/bioengineering10050525"},{"key":"e_1_2_8_36_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2023.121482"},{"key":"e_1_2_8_37_1","unstructured":"LehtinenJ.Noise2Noise: learning image restoration without clean data.arXiv.2018. arXiv:1803.04189."},{"key":"e_1_2_8_38_1","volume-title":"Diagnostic Ultrasound Imaging: Inside Out","author":"Szabo TL","year":"2014"},{"key":"e_1_2_8_39_1","doi-asserted-by":"crossref","unstructured":"LiebgottH Rodriguez\u2010MolaresA CervenanskyF JensenJA BernardO.Bernard O. Plane\u2010wave imaging challenge in medical ultrasound. In: 2016 IEEE International Ultrasonics Symposium (IUS).IEEE2016:1\u20104.","DOI":"10.1109\/ULTSYM.2016.7728908"},{"key":"e_1_2_8_40_1","doi-asserted-by":"publisher","DOI":"10.1109\/TUFFC.2021.3094849"},{"key":"e_1_2_8_41_1","doi-asserted-by":"crossref","unstructured":"RonnebergerO FischerP BroxT U\u2010Net: convolutional networks for biomedical image segmentation.International Conference on Medical Image Computing and Computer\u2010Assisted Intervention.2015:234\u2010241.","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"e_1_2_8_42_1","doi-asserted-by":"crossref","unstructured":"JacobB KligysS ChenB et\u00a0al.Quantization and training of neural networks for efficient integer\u2010arithmetic\u2010only inference.Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2018. doi:10.1109\/CVPR.2018.00286","DOI":"10.1109\/CVPR.2018.00286"},{"key":"e_1_2_8_43_1","unstructured":"Advanced Micro Devices Inc.Vitis AI User Guide (UG1414). Version 3.5. Advanced Micro Devices Inc.;2023."},{"key":"e_1_2_8_44_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.dib.2019.104863"},{"issue":"8","key":"e_1_2_8_45_1","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0200412","article-title":"Automated measurement of fetal head circumference using 2D ultrasound images","volume":"13","author":"Heuvel TL","year":"2018","journal-title":"PLoS One"},{"key":"e_1_2_8_46_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41746\u2010022\u201000592\u2010y"},{"key":"e_1_2_8_47_1","doi-asserted-by":"crossref","unstructured":"TobinJ FongR RayA SchneiderJ ZarembaW AbbeelP Domain randomization for transferring deep neural networks from simulation to the real world. In:2017 IEEE\/RSJ International Conference on Intelligent Robots and Systems (IROS).IEEE;2017:23\u201030. doi:10.1109\/IROS.2017.8202133","DOI":"10.1109\/IROS.2017.8202133"},{"key":"e_1_2_8_48_1","unstructured":"LoshchilovI.Decoupled weight decay regularization.arXiv.2017. arXiv:1711.05101."},{"key":"e_1_2_8_49_1","doi-asserted-by":"publisher","DOI":"10.1109\/TBME.2015.2486042"},{"key":"e_1_2_8_50_1","doi-asserted-by":"publisher","DOI":"10.2307\/3001968"},{"issue":"2","key":"e_1_2_8_51_1","first-page":"65","article-title":"A simple sequentially rejective multiple test procedure","volume":"6","author":"Holm S","year":"1979","journal-title":"Scand J Stat"},{"key":"e_1_2_8_52_1","doi-asserted-by":"crossref","unstructured":"KerbyDS.The simple difference formula: an approach to teaching nonparametric correlation.Compr Psychol.2014;3:11.IT.3.1.10.2466\/11.IT.3.1","DOI":"10.2466\/11.IT.3.1"},{"key":"e_1_2_8_53_1","doi-asserted-by":"publisher","DOI":"10.3102\/10769986006002107"}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.70620","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/full-xml\/10.1002\/mp.70620","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.70620","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T13:42:34Z","timestamp":1788270154000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.70620"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8]]},"references-count":52,"journal-issue":{"issue":"8","published-print":{"date-parts":[[2026,8]]}},"alternative-id":["10.1002\/mp.70620"],"URL":"https:\/\/doi.org\/10.1002\/mp.70620","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8]]},"assertion":[{"value":"2026-01-20","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-07-17","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-08-03","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e70620"}}