{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,5]],"date-time":"2026-11-05T04:46:36Z","timestamp":1793853996848,"version":"4.6.1"},"reference-count":33,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T00:00:00Z","timestamp":1750291200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T00:00:00Z","timestamp":1750291200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"Sakari Alhopuron s\u00e4\u00e4ti\u00f6"},{"DOI":"10.13039\/501100006306","name":"Sigrid Jus\u00e9liuksen S\u00e4\u00e4ti\u00f6","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100006306","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Eur Radiol Exp"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>In acute neck infections, magnetic resonance imaging (MRI) shows retropharyngeal edema (RPE), which is a prognostic imaging biomarker for a severe course of illness. This study aimed to develop a deep learning-based algorithm for the automated detection of RPE.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We developed a deep neural network consisting of two parts using axial T2-weighted water-only Dixon MRI images from 479 patients with acute neck infections annotated by radiologists at both slice and patient levels. First, a convolutional neural network (CNN) classified individual slices; second, an algorithm classified patients based on a stack of slices. Model performance was compared with the radiologists\u2019 assessment as a reference standard. Accuracy, sensitivity, specificity, and area under receiver operating characteristic curve (AUROC) were calculated. The proposed CNN was compared with InceptionV3, and the patient-level classification algorithm was compared with traditional machine learning models.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Of the 479 patients, 244 (51%) were positive and 235 (49%) negative for RPE. Our model achieved accuracy, sensitivity, specificity, and AUROC of 94.6%, 83.3%, 96.2%, and 94.1% at the slice level, and 87.4%, 86.5%, 88.2%, and 94.8% at the patient level, respectively. The proposed CNN was faster than InceptionV3 but equally accurate. Our patient classification algorithm outperformed traditional machine learning models.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>A deep learning model, based on weakly annotated data and computationally manageable training, achieved high accuracy for automatically detecting RPE on MRI in patients with acute neck infections.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Relevance statement<\/jats:title>\n                    <jats:p>Our automated method for detecting relevant MRI findings was efficiently trained and might be easily deployed in practice to study clinical applicability. This approach might improve early detection of patients at high risk for a severe course of acute neck infections.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Key Points<\/jats:title>\n                    <jats:p>\n                      <jats:list list-type=\"bullet\">\n                        <jats:list-item>\n                          <jats:p>Deep learning automatically detected retropharyngeal edema on MRI in acute neck infections.<\/jats:p>\n                        <\/jats:list-item>\n                        <jats:list-item>\n                          <jats:p>Areas under the receiver operating characteristic curve were 94.1% at the slice level and 94.8% at the patient level.<\/jats:p>\n                        <\/jats:list-item>\n                        <jats:list-item>\n                          <jats:p>The proposed convolutional neural network was lightweight and required only weakly annotated data.<\/jats:p>\n                        <\/jats:list-item>\n                      <\/jats:list>\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Graphical Abstract<\/jats:title>\n                  <\/jats:sec>","DOI":"10.1186\/s41747-025-00599-6","type":"journal-article","created":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T10:19:16Z","timestamp":1750328356000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":4,"title":["Deep learning detects retropharyngeal edema on MRI in patients with acute neck infections"],"prefix":"10.1186","volume":"9","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7775-7656","authenticated-orcid":false,"given":"Oona","family":"Rainio","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1581-1709","authenticated-orcid":false,"given":"Heidi","family":"Huhtanen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-5933-5421","authenticated-orcid":false,"given":"Jari-Pekka","family":"Vierula","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2482-9051","authenticated-orcid":false,"given":"Janne","family":"Nurminen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8007-5090","authenticated-orcid":false,"given":"Jaakko","family":"Heikkinen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2104-6302","authenticated-orcid":false,"given":"Mikko","family":"Nyman","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0982-8360","authenticated-orcid":false,"given":"Riku","family":"Kl\u00e9n","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0368-3746","authenticated-orcid":false,"given":"Jussi","family":"Hirvonen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,6,19]]},"reference":[{"key":"599_CR1","doi-asserted-by":"publisher","first-page":"1241","DOI":"10.1007\/s00405-011-1761-1","volume":"269","author":"P Boscolo-Rizzo","year":"2012","unstructured":"Boscolo-Rizzo P, Stellin M, Muzzi E et al (2012) Deep neck infections: a study of 365 cases highlighting recommendations for management and treatment. Eur Arch Otorhinolaryngol 269:1241\u20131249. https:\/\/doi.org\/10.1007\/s00405-011-1761-1","journal-title":"Eur Arch Otorhinolaryngol"},{"key":"599_CR2","doi-asserted-by":"publisher","first-page":"1808","DOI":"10.1148\/rg.2019190159","volume":"39","author":"S Kamalian","year":"2019","unstructured":"Kamalian S, Avery L, Lev MH et al (2019) Nontraumatic head and neck emergencies. Radiographics 39:1808\u20131823. https:\/\/doi.org\/10.1148\/rg.2019190159","journal-title":"Radiographics"},{"key":"599_CR3","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1186\/s13244-022-01347-9","volume":"14","author":"J Hirvonen","year":"2023","unstructured":"Hirvonen J, Heikkinen J, Nyman M et al (2023) MRI of acute neck infections: evidence summary and pictorial review. Insights Imaging 14:5. https:\/\/doi.org\/10.1186\/s13244-022-01347-9","journal-title":"Insights Imaging"},{"key":"599_CR4","doi-asserted-by":"publisher","first-page":"735","DOI":"10.1177\/0284185120940242","volume":"62","author":"J Nurminen","year":"2021","unstructured":"Nurminen J, Velhonoja J, Heikkinen J et al (2021) Emergency neck MRI: feasibility and diagnostic accuracy in cases of neck infection. Acta Radiologica 62:735\u2013742. https:\/\/doi.org\/10.1177\/0284185120940242","journal-title":"Acta Radiologica"},{"key":"599_CR5","doi-asserted-by":"publisher","first-page":"1158","DOI":"10.1007\/s00247-021-05275-6","volume":"52","author":"J Nurminen","year":"2022","unstructured":"Nurminen J, Heikkinen J, Happonen T et al (2022) Magnetic resonance imaging findings in pediatric neck infections\u2014a comparison with adult patients. Pediatr Radiol 52:1158\u20131166. https:\/\/doi.org\/10.1007\/s00247-021-05275-6","journal-title":"Pediatr Radiol"},{"key":"599_CR6","doi-asserted-by":"publisher","first-page":"1078","DOI":"10.1007\/s00330-021-08200-5","volume":"32","author":"J Heikkinen","year":"2022","unstructured":"Heikkinen J, Nurminen J, Velhonoja J et al (2022) Clinical and prognostic significance of emergency MRI findings in neck infections. Eur Radiol 32:1078\u20131086. https:\/\/doi.org\/10.1007\/s00330-021-08200-5","journal-title":"Eur Radiol"},{"key":"599_CR7","doi-asserted-by":"publisher","first-page":"286","DOI":"10.3174\/ajnr.A7368","volume":"43","author":"J Heikkinen","year":"2022","unstructured":"Heikkinen J, Nurminen J, Velhonoja J et al (2022) MRI findings in acute tonsillar infections. AJNR Am J Neuroradiol 43:286\u2013291. https:\/\/doi.org\/10.3174\/ajnr.A7368","journal-title":"AJNR Am J Neuroradiol"},{"key":"599_CR8","doi-asserted-by":"publisher","first-page":"364","DOI":"10.1007\/s11282-022-00646-7","volume":"39","author":"J Heikkinen","year":"2023","unstructured":"Heikkinen J, Jokihaka V, Nurminen J et al (2023) MRI of odontogenic maxillofacial infections: diagnostic accuracy and reliability. Oral Radiol 39:364\u2013371. https:\/\/doi.org\/10.1007\/s11282-022-00646-7","journal-title":"Oral Radiol"},{"key":"599_CR9","doi-asserted-by":"publisher","first-page":"65","DOI":"10.1186\/s41747-023-00379-0","volume":"7","author":"JP Vierula","year":"2023","unstructured":"Vierula JP, Nurminen J, Jussila V et al (2023) Diagnostic performance of short noncontrast biparametric 3-T MRI for tonsillar infections: comparison with a full protocol including contrast-enhanced sequences. Eur Radiol Exp 7:65. https:\/\/doi.org\/10.1186\/s41747-023-00379-0","journal-title":"Eur Radiol Exp"},{"key":"599_CR10","doi-asserted-by":"publisher","first-page":"22","DOI":"10.1186\/s13244-019-0832-5","volume":"11","author":"E Montagnon","year":"2020","unstructured":"Montagnon E, Cerny M, Cadrin-Ch\u00eanevert A et al (2020) Deep learning workflow in radiology: a primer. Insights Imaging 11:22. https:\/\/doi.org\/10.1186\/s13244-019-0832-5","journal-title":"Insights Imaging"},{"key":"599_CR11","doi-asserted-by":"publisher","DOI":"10.1186\/s12880-022-00763-z","volume":"22","author":"H Huhtanen","year":"2022","unstructured":"Huhtanen H, Nyman M, Mohsen T et al (2022) Automated detection of pulmonary embolism from CT-angiograms using deep learning. BMC Med Imaging 22:43. https:\/\/doi.org\/10.1186\/s12880-022-00763-z","journal-title":"BMC Med Imaging"},{"key":"599_CR12","doi-asserted-by":"publisher","first-page":"655","DOI":"10.1148\/radiol.212526","volume":"305","author":"KK Bressem","year":"2022","unstructured":"Bressem KK, Adams LC, Proft F et al (2022) Deep learning detects changes indicative of axial spondyloarthritis at MRI of sacroiliac joints. Radiology 305:655\u2013665. https:\/\/doi.org\/10.1148\/radiol.212526","journal-title":"Radiology"},{"key":"599_CR13","doi-asserted-by":"publisher","first-page":"625","DOI":"10.1007\/s00296-020-04511-w","volume":"40","author":"I Kucybala","year":"2020","unstructured":"Kucybala I, Tabor Z, Polak J et al (2020) The semiautomated algorithm for the detection of bone marrow oedema lesions in patients with axial spondyloarthritis. Rheumatol Int 40:625\u2013633. https:\/\/doi.org\/10.1007\/s00296-020-04511-w","journal-title":"Rheumatol Int"},{"key":"599_CR14","doi-asserted-by":"publisher","first-page":"4198","DOI":"10.1093\/rheumatology\/keac059","volume":"61","author":"KYY Lin","year":"2022","unstructured":"Lin KYY, Peng C, Lee KH et al (2022) Deep learning algorithms for magnetic resonance imaging of inflammatory sacroiliitis in axial spondyloarthritis. Rheumatology 61:4198\u20134206. https:\/\/doi.org\/10.1093\/rheumatology\/keac059","journal-title":"Rheumatology"},{"key":"599_CR15","doi-asserted-by":"publisher","first-page":"1156","DOI":"10.3390\/diagnostics11071156","volume":"11","author":"KH Lee","year":"2021","unstructured":"Lee KH, Choi ST, Lee GY et al (2021) Method for diagnosing the bone marrow edema of sacroiliac joint in patients with axial spondyloarthritis using magnetic resonance image analysis based on deep learning. Diagnostics (Basel) 11:1156. https:\/\/doi.org\/10.3390\/diagnostics11071156","journal-title":"Diagnostics (Basel)"},{"key":"599_CR16","doi-asserted-by":"publisher","first-page":"4852","DOI":"10.3390\/jcm12144852","volume":"12","author":"J Ozga","year":"2023","unstructured":"Ozga J, Wyka M, Raczko A et al (2023) Performance of fully automated algorithm detecting bone marrow edema in sacroiliac joints. J Clin Med 12:4852. https:\/\/doi.org\/10.3390\/jcm12144852","journal-title":"J Clin Med"},{"key":"599_CR17","doi-asserted-by":"publisher","DOI":"10.3389\/fphys.2023.1132214","volume":"14","author":"Y Zheng","year":"2023","unstructured":"Zheng Y, Bai C, Zheng K et al (2023) Deep-learning based quantification model for hip bone marrow edema and synovitis in patients with spondyloarthritis based on magnetic resonance images. Front Physiol 14:1132214. https:\/\/doi.org\/10.3389\/fphys.2023.1132214","journal-title":"Front Physiol"},{"key":"599_CR18","doi-asserted-by":"publisher","first-page":"526","DOI":"10.3390\/bioengineering11060526","volume":"11","author":"AS Yu","year":"2024","unstructured":"Yu AS, Yang M, Lartey R et al (2024) Unsupervised segmentation of knee bone marrow edema-like lesions using conditional generative models. Bioengineering 11:526. https:\/\/doi.org\/10.3390\/bioengineering11060526","journal-title":"Bioengineering"},{"key":"599_CR19","doi-asserted-by":"publisher","first-page":"4125","DOI":"10.1007\/s00586-023-08099-0","volume":"33","author":"Y Lin","year":"2024","unstructured":"Lin Y, Chan SCW, Chung HY et al (2024) A deep neural network for MRI spinal inflammation in axial spondyloarthritis. Eur Spine J 33:4125\u20134134. https:\/\/doi.org\/10.1007\/s00586-023-08099-0","journal-title":"Eur Spine J"},{"key":"599_CR20","doi-asserted-by":"publisher","first-page":"1885","DOI":"10.1007\/s10278-023-00812-1","volume":"36","author":"O Rainio","year":"2023","unstructured":"Rainio O, Han C, Teuho J et al (2023) Carimas: an extensive medical imaging data processing yool for research. J Digit Imaging 36:1885\u20131893. https:\/\/doi.org\/10.1007\/s10278-023-00812-1","journal-title":"J Digit Imaging"},{"key":"599_CR21","unstructured":"Van Rossum G, Drake FL (2009) Python 3 Reference Manual ISBN: 1441412697. CreateSpace, Scotts Valley, CA"},{"key":"599_CR22","unstructured":"Chollet F (2015) Keras https:\/\/github.com\/fchollet\/keras"},{"key":"599_CR23","unstructured":"Abadi M, Agarwal A, Barham P et al (2015) TensorFlow: large-scale machine learning on heterogeneous systems. Mountain View, CA: Tensorflow"},{"key":"599_CR24","first-page":"2825","volume":"12","author":"F Pedregosa","year":"2011","unstructured":"Pedregosa F, Varoquaux G, Gramfort A (2011) Scikit-learn: machine learning in Python. J Mach Learn Res 12:2825\u20132830","journal-title":"J Mach Learn Res"},{"key":"599_CR25","doi-asserted-by":"publisher","first-page":"261","DOI":"10.1038\/s41592-019-0686-2","volume":"17","author":"P Virtanen","year":"2020","unstructured":"Virtanen P, Gommers R, Oliphant TE et al (2020) SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat Methods 17:261\u2013272. https:\/\/doi.org\/10.1038\/s41592-019-0686-2","journal-title":"Nat Methods"},{"key":"599_CR26","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-023-37603-1","volume":"13","author":"H Hellstr\u00f6m","year":"2023","unstructured":"Hellstr\u00f6m H, Liedes J, Rainio O et al (2023) Classification of head and neck cancer from PET images using convolutional neural networks. Sci Rep 13:10528. https:\/\/doi.org\/10.1038\/s41598-023-37603-1","journal-title":"Sci Rep"},{"key":"599_CR27","doi-asserted-by":"crossref","unstructured":"Ronneberger O, Fischer P, Brox T (2015) U-Net: Convolutional Networks for Biomedical Image Segmentation in Medical Image Computing and Computer-Assisted Intervention\u2014MICCAI 2015: 18th International Conference, Munich, Germany, October 5\u20139, Proceedings, Part III vol. 18. pp. 234\u2013241","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"599_CR28","doi-asserted-by":"publisher","first-page":"92","DOI":"10.1016\/j.neucom.2020.04.157","volume":"444","author":"H Yu","year":"2021","unstructured":"Yu H, Yang LT, Zhang Q et al (2021) Convolutional neural networks for medical image analysis: state-of-the-art, comparisons, improvement and perspectives. Neurocomputing 444:92\u2013110. https:\/\/doi.org\/10.1016\/j.neucom.2020.04.157","journal-title":"Neurocomputing"},{"key":"599_CR29","doi-asserted-by":"crossref","unstructured":"Szegedy C, Vanhoucke V, Ioffe S et al (2016) Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. IEEE. pp. 2818\u20132826","DOI":"10.1109\/CVPR.2016.308"},{"key":"599_CR30","doi-asserted-by":"publisher","first-page":"6999","DOI":"10.1109\/TNNLS.2021.3084827","volume":"33","author":"Z Li","year":"2021","unstructured":"Li Z, Liu F, Yang W et al (2021) A survey of convolutional neural networks: analysis, applications, and prospects. IEEE Trans Neural Netw Learn Syst 33:6999\u20137019. https:\/\/doi.org\/10.1109\/TNNLS.2021.3084827","journal-title":"IEEE Trans Neural Netw Learn Syst"},{"key":"599_CR31","doi-asserted-by":"publisher","first-page":"3353","DOI":"10.1007\/s11760-024-02998-5","volume":"18","author":"O Rainio","year":"2024","unstructured":"Rainio O, Kl\u00e9n R (2024) Comparison of simple augmentation transformations for a convolutional neural network classifying medical images. SIViP 18:3353\u20133360. https:\/\/doi.org\/10.1007\/s11760-024-02998-5","journal-title":"SIViP"},{"key":"599_CR32","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-024-56706-x","volume":"14","author":"O Rainio","year":"2024","unstructured":"Rainio O, Teuho J, Kl\u00e9n R (2024) Evaluation metrics and statistical tests for machine learning. Sci Rep 14:6086. https:\/\/doi.org\/10.1038\/s41598-024-56706-x","journal-title":"Sci Rep"},{"key":"599_CR33","doi-asserted-by":"publisher","unstructured":"Rainio O, Tamminen J, Ven\u00e4l\u00e4inen MS et al (2024) Comparison of thresholds for a convolutional neural network classifying medical images. Int J Data Sci Anal https:\/\/doi.org\/10.1007\/s41060-024-00584-z","DOI":"10.1007\/s41060-024-00584-z"}],"container-title":["European Radiology Experimental"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s41747-025-00599-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s41747-025-00599-6\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s41747-025-00599-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T10:19:18Z","timestamp":1750328358000},"score":1,"resource":{"primary":{"URL":"https:\/\/eurradiolexp.springeropen.com\/articles\/10.1186\/s41747-025-00599-6"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,19]]},"references-count":33,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,12]]}},"alternative-id":["599"],"URL":"https:\/\/doi.org\/10.1186\/s41747-025-00599-6","relation":{},"ISSN":["2509-9280"],"issn-type":[{"value":"2509-9280","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,6,19]]},"assertion":[{"value":"24 January 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"4 June 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"19 June 2025","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"According to the national legislation, no separate ethics committee approval is needed for retrospective studies that involve a secondary use of registry or archival data.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval and consent to participate"}},{"value":"According to the national legislation, no separate ethics committee approval is needed for retrospective studies that involve a secondary use of registry or archival data.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}},{"value":"The authors declare that they have no competing interests.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"60"}}