{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,18]],"date-time":"2026-03-18T09:45:11Z","timestamp":1773827111794,"version":"3.50.1"},"reference-count":35,"publisher":"MDPI AG","issue":"8","license":[{"start":{"date-parts":[[2022,8,19]],"date-time":"2022-08-19T00:00:00Z","timestamp":1660867200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["J. Imaging"],"abstract":"<jats:p>A variety of medical computer vision applications analyze 2D slices of computed tomography (CT) scans, whereas axial slices from the body trunk region are usually identified based on their relative position to the spine. A limitation of such systems is that either the correct slices must be extracted manually or labels of the vertebrae are required for each CT scan to develop an automated extraction system. In this paper, we propose an unsupervised domain adaptation (UDA) approach for vertebrae detection and identification based on a novel Domain Sanity Loss (DSL) function. With UDA the model\u2019s knowledge learned on a publicly available (source) data set can be transferred to the target domain without using target labels, where the target domain is defined by the specific setup (CT modality, study protocols, applied pre- and processing) at the point of use (e.g., a specific clinic with its specific CT study protocols). With our approach, a model is trained on the source and target data set in parallel. The model optimizes a supervised loss for labeled samples from the source domain and the DSL loss function based on domain-specific \u201csanity checks\u201d for samples from the unlabeled target domain. Without using labels from the target domain, we are able to identify vertebra centroids with an accuracy of 72.8%. By adding only ten target labels during training the accuracy increases to 89.2%, which is on par with the current state-of-the-art for full supervised learning, while using about 20 times less labels. Thus, our model can be used to extract 2D slices from 3D CT scans on arbitrary data sets fully automatically without requiring an extensive labeling effort, contributing to the clinical adoption of medical imaging by hospitals.<\/jats:p>","DOI":"10.3390\/jimaging8080222","type":"journal-article","created":{"date-parts":[[2022,8,21]],"date-time":"2022-08-21T23:51:01Z","timestamp":1661125861000},"page":"222","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":7,"title":["Unsupervised Domain Adaptation for Vertebrae Detection and Identification in 3D CT Volumes Using a Domain Sanity Loss"],"prefix":"10.3390","volume":"8","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-8084-2317","authenticated-orcid":false,"given":"Pascal","family":"Sager","sequence":"first","affiliation":[{"name":"Centre for AI, Technikumstrasse 71, Zurich University of Applied Sciences, 8400 Winterthur, Switzerland"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Sebastian","family":"Salzmann","sequence":"additional","affiliation":[{"name":"Centre for AI, Technikumstrasse 71, Zurich University of Applied Sciences, 8400 Winterthur, Switzerland"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Felice","family":"Burn","sequence":"additional","affiliation":[{"name":"Cantonal Hospital Aarau, AI and Data Science CoE, Tellstrasse 25, 5001 Aarau, Switzerland"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3784-0420","authenticated-orcid":false,"given":"Thilo","family":"Stadelmann","sequence":"additional","affiliation":[{"name":"Centre for AI, Technikumstrasse 71, Zurich University of Applied Sciences, 8400 Winterthur, Switzerland"},{"name":"ECLT European Centre for Living Technology, 30123 Venice, Italy"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2022,8,19]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","unstructured":"Esmaeilzadeh, P. (2020). Use of AI-based tools for healthcare purposes: A survey study from consumer\u2019s perspectives. BMC Med Informatics Decis. Mak., 20.","DOI":"10.1186\/s12911-020-01191-1"},{"key":"ref_2","first-page":"1","article-title":"Body Composition Analysis of Computed Tomography Scans in Clinical Populations: The Role of Deep Learning","volume":"13","author":"Paris","year":"2019","journal-title":"Lifestyle Genom."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"60","DOI":"10.1016\/j.media.2017.07.005","article-title":"A survey on deep learning in medical image analysis","volume":"42","author":"Litjens","year":"2017","journal-title":"Med. Image Anal."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"1795","DOI":"10.1007\/s00330-020-07147-3","article-title":"Fully automated body composition analysis in routine CT imaging using 3D semantic segmentation convolutional neural networks","volume":"31","author":"Koitka","year":"2021","journal-title":"Eur. Radiol."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"5723","DOI":"10.1002\/mp.14465","article-title":"Automatic segmentation of CT images for ventral body composition analysis","volume":"47","author":"Fu","year":"2020","journal-title":"Med. Phys."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"669","DOI":"10.1148\/radiol.2018181432","article-title":"Automated Abdominal Segmentation of CT Scans for Body Composition Analysis Using Deep Learning","volume":"290","author":"Weston","year":"2019","journal-title":"Radiology"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"2143","DOI":"10.3945\/jn.116.236844","article-title":"Estimation of Skeletal Muscle Mass and Visceral Adipose Tissue Volume by a Single Magnetic Resonance Imaging Slice in Healthy Elderly Adults","volume":"146","author":"Schweitzer","year":"2016","journal-title":"J. Nutr."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"109943","DOI":"10.1016\/j.ejrad.2021.109943","article-title":"Methodology, clinical applications, and future directions of body composition analysis using computed tomography (CT) images: A review","volume":"145","author":"Tolonen","year":"2021","journal-title":"Eur. J. Radiol."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"2333","DOI":"10.1152\/japplphysiol.00744.2004","article-title":"Total body skeletal muscle and adipose tissue volumes: Estimation from a single abdominal cross-sectional image","volume":"97","author":"Shen","year":"2004","journal-title":"J. Appl. Physiol."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"512","DOI":"10.1109\/TMI.2015.2479252","article-title":"Body Composition Assessment in Axial CT Images Using FEM-Based Automatic Segmentation of Skeletal Muscle","volume":"35","author":"Popuri","year":"2016","journal-title":"IEEE Trans. Med Imaging"},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"3142","DOI":"10.1007\/s00330-021-08313-x","article-title":"End-to-end automated body composition analyses with integrated quality control for opportunistic assessment of sarcopenia in CT","volume":"32","author":"Nowak","year":"2021","journal-title":"Eur. Radiol."},{"key":"ref_12","first-page":"1","article-title":"Data Centrism and the Core of Data Science as a Scientific Discipline","volume":"8","author":"Stadelmann","year":"2022","journal-title":"Arch. Data Sci. Ser. A (Online First)"},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Simmler, N., Sager, P., Andermatt, P., Chavarriaga, R., Schilling, F.P., Rosenthal, M., and Stadelmann, T. (2021, January 9). A Survey of Un-, Weakly-, and Semi-Supervised Learning Methods for Noisy, Missing and Partial Labels in Industrial Vision Applications. Proceedings of the 8th Swiss Conference on Data Science (SDS), Lucerne, Switzerland.","DOI":"10.1109\/SDS51136.2021.00012"},{"key":"ref_14","unstructured":"McCouat, J., and Glocker, B. (2019). Vertebrae Detection and Localization in CT with Two-Stage CNNs and Dense Annotations. arXiv."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"513","DOI":"10.1007\/s10278-018-0130-7","article-title":"Automatic Lumbar MRI Detection and Identification Based on Deep Learning","volume":"32","author":"Zhou","year":"2019","journal-title":"J. Digit. Imaging"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Yi, J., Wu, P., Huang, Q., Qu, H., and Metaxas, D.N. (2020, January 3\u20137). Vertebra-Focused Landmark Detection for Scoliosis Assessment. Proceedings of the 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI), Iowa City, IA, USA.","DOI":"10.1109\/ISBI45749.2020.9098675"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Altini, N., De Giosa, G., Fragasso, N., Coscia, C., Sibilano, E., Prencipe, B., Hussain, S.M., Brunetti, A., Buongiorno, D., and Guerriero, A. (2021). Segmentation and Identification of Vertebrae in CT Scans Using CNN, k-Means Clustering and k-NN. Informatics, 8.","DOI":"10.3390\/informatics8020040"},{"key":"ref_18","first-page":"590","article-title":"Automatic Localization and Identification of Vertebrae in Arbitrary Field-of-View CT Scans","volume":"Volume 7512","author":"Hutchison","year":"2012","journal-title":"Proceedings of the Medical Image Computing and Computer-Assisted Intervention\u2014MICCAI"},{"key":"ref_19","first-page":"262","article-title":"Vertebrae Localization in Pathological Spine CT via Dense Classification from Sparse Annotations","volume":"Volume 7908","author":"Hutchison","year":"2013","journal-title":"Proceedings of the Medical Image Computing and Computer-Assisted Intervention\u2014MICCAI"},{"key":"ref_20","unstructured":"Niethammer, M., Styner, M., Aylward, S., Zhu, H., Oguz, I., Yap, P.T., and Shen, D. (2017, January 25\u201330). Automatic Vertebra Labeling in Large-Scale 3D CT using Deep Image-to-Image Network with Message Passing and Sparsity Regularization. Proceedings of the International Conference on Information Processing in Medical Imaging, Boone, NC, USA."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"1266","DOI":"10.1109\/TMI.2018.2798293","article-title":"Joint Vertebrae Identification and Localization in Spinal CT Images by Combining Short- and Long-Range Contextual Information","volume":"37","author":"Liao","year":"2018","journal-title":"IEEE Trans. Med. Imaging"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Navab, N., Hornegger, J., Wells, W.M., and Frangi, A.F. (2015, January 5\u20139). U-Net: Convolutional Networks for Biomedical Image Segmentation. Proceedings of the Medical Image Computing and Computer-Assisted Intervention\u2014MICCAI, Munich, Germany.","DOI":"10.1007\/978-3-319-24571-3"},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Stadelmann, T., Tolkachev, V., Sick, B., Stampfli, J., and D\u00fcrr, O. (2019). Beyond ImageNet: Deep learning in industrial practice. Applied Data Science: Lessons Learned for the Data-Driven Business, Springer International Publishing.","DOI":"10.1007\/978-3-030-11821-1_12"},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Amirian, M., Montoya-Zegarra, J.A., Gruss, J., Stebler, Y.D., Bozkir, A.S., Calandri, M., Schwenker, F., and Stadelmann, T. (2021, January 23\u201325). PrepNet: A Convolutional Auto-Encoder to Homogenize CT Scans for Cross-Dataset Medical Image Analysis. Proceedings of the 14th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI), Shanghai, China.","DOI":"10.1109\/CISP-BMEI53629.2021.9624344"},{"key":"ref_25","unstructured":"Kingma, D.P., and Ba, J. (2014). Adam: A Method for Stochastic Optimization. arXiv."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"167","DOI":"10.1023\/B:VISI.0000022288.19776.77","article-title":"Efficient Graph-Based Image Segmentation","volume":"59","author":"Felzenszwalb","year":"2004","journal-title":"Int. J. Comput. Vis."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"1104","DOI":"10.1007\/s00586-010-1326-9","article-title":"Comparative anatomical dimensions of the complete human and porcine spine","volume":"19","author":"Busscher","year":"2010","journal-title":"Eur. Spine J."},{"key":"ref_28","unstructured":"Biomedical Image Analysis Group, Imperial College London (2021, November 15). BioMedIA Spine Dataset. Available online: https:\/\/biomedia.doc.ic.ac.uk\/data\/spine\/."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s13104-021-05592-x","article-title":"COVID19-CT-dataset: An open-access chest CT image repository of 1000+ patients with confirmed COVID-19 diagnosis","volume":"14","author":"Shakouri","year":"2021","journal-title":"BMC Res. Notes"},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"178","DOI":"10.1186\/s13104-021-05592-x","article-title":"COVID19-CT-Dataset: An Open-Access Chest CT Image Repository of 1000+ Patients with Confirmed COVID-19 Diagnosis","volume":"14","author":"Mostafavi","year":"2021","journal-title":"BMC Res. Notes"},{"key":"ref_31","first-page":"515","article-title":"Automatic Localization and Identification of Vertebrae in Spine CT via a Joint Learning Model with Deep Neural Networks","volume":"Volume 9349","author":"Navab","year":"2015","journal-title":"Proceedings of the Medical Image Computing and Computer-Assisted Intervention\u2014MICCAI"},{"key":"ref_32","unstructured":"Ma, D., Chow, V., Popuri, K., and Beg, M.F. (2021). Comprehensive Validation of Automated Whole Body Skeletal Muscle, Adipose Tissue, and Bone Segmentation from 3D CT images for Body Composition Analysis: Towards Extended Body Composition. arXiv."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"280","DOI":"10.1016\/j.media.2019.03.009","article-title":"Not-so-supervised: A survey of semi-supervised, multi-instance, and transfer learning in medical image analysis","volume":"54","author":"Cheplygina","year":"2019","journal-title":"Med Image Anal."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"1173","DOI":"10.1109\/TBME.2021.3117407","article-title":"Domain Adaptation for Medical Image Analysis: A Survey","volume":"69","author":"Guan","year":"2022","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Tuggener, L., Satyawan, Y.P., Pacha, A., Schmidhuber, J., and Stadelmann, T. (2021, January 10\u201315). The DeepScoresV2 dataset and benchmark for music object detection. Proceedings of the 25th International Conference on Pattern Recognition (ICPR), Milan, Italy.","DOI":"10.1109\/ICPR48806.2021.9412290"}],"container-title":["Journal of Imaging"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2313-433X\/8\/8\/222\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T00:12:17Z","timestamp":1760141537000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2313-433X\/8\/8\/222"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,8,19]]},"references-count":35,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2022,8]]}},"alternative-id":["jimaging8080222"],"URL":"https:\/\/doi.org\/10.3390\/jimaging8080222","relation":{},"ISSN":["2313-433X"],"issn-type":[{"value":"2313-433X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,8,19]]}}}