{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,21]],"date-time":"2026-10-21T16:30:03Z","timestamp":1792600203228,"version":"4.5.0"},"reference-count":56,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,12,11]],"date-time":"2024-12-11T00:00:00Z","timestamp":1733875200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Radiol."],"abstract":"<jats:p>Pneumothorax, a life-threatening condition characterized by air accumulation in the pleural cavity, requires early and accurate detection for optimal patient outcomes. Chest X-ray radiographs are a common diagnostic tool due to their speed and affordability. However, detecting pneumothorax can be challenging for radiologists because the sole visual indicator is often a thin displaced pleural line. This research explores deep learning techniques to automate and improve the detection and segmentation of pneumothorax from chest X-ray radiographs. We propose a novel architecture that combines the advantages of fully convolutional neural networks (FCNNs) and Vision Transformers (ViTs) while using only convolutional modules to avoid the quadratic complexity of ViT\u2019s self-attention mechanism. This architecture utilizes a patch-based encoder-decoder structure with skip connections to effectively combine high-level and low-level features. Compared to prior research and baseline FCNNs, our model demonstrates significantly higher accuracy in detection and segmentation while maintaining computational efficiency. This is evident on two datasets: (1) the SIIM-ACR Pneumothorax Segmentation dataset and (2) a novel dataset we curated from The Medical City, a private hospital in the Philippines. Ablation studies further reveal that using a mixed Tversky and Focal loss function significantly improves performance compared to using solely the Tversky loss. Our findings suggest our model has the potential to improve diagnostic accuracy and efficiency in pneumothorax detection, potentially aiding radiologists in clinical settings.<\/jats:p>","DOI":"10.3389\/fradi.2024.1424065","type":"journal-article","created":{"date-parts":[[2024,12,11]],"date-time":"2024-12-11T04:34:16Z","timestamp":1733891656000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["Pneumothorax detection and segmentation from chest X-ray radiographs using a patch-based fully convolutional encoder-decoder network"],"prefix":"10.3389","volume":"4","author":[{"given":"Jakov Ivan S.","family":"Dumbrique","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Reynan B.","family":"Hernandez","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Juan Miguel L.","family":"Cruz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ryan M.","family":"Pagdanganan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"suffix":"Jr","given":"Prospero C.","family":"Naval","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2024,12,11]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"103458","DOI":"10.1016\/j.bspc.2021.103458","article-title":"Semantic segmentation of pancreatic medical images by using convolutional neural network","volume":"73","author":"Huang","year":"2022","journal-title":"Biomed Signal Process Control"},{"key":"B2","doi-asserted-by":"publisher","first-page":"94631","DOI":"10.1109\/ACCESS.2020.2995567","article-title":"Cardiomegaly detection on chest radiographs: segmentation versus classification","volume":"8","author":"Sogancioglu","year":"2020","journal-title":"IEEE Access"},{"key":"B3","doi-asserted-by":"publisher","first-page":"605132","DOI":"10.3389\/fbioe.2020.605132","article-title":"Ra-unet: a hybrid deep attention-aware network to extract liver and tumor in ct scans","volume":"8","author":"Jin","year":"2020","journal-title":"Front Bioeng Biotechnol"},{"key":"B4","doi-asserted-by":"crossref","first-page":"205","DOI":"10.1117\/12.2628519","article-title":"Ma-unet: an improved version of unet based on multi-scale and attention mechanism for medical image segmentation","volume-title":"Third International Conference on Electronics and Communication; Network and Computer Technology (ECNCT 2021)","author":"Cai","year":"2022"},{"key":"B5","article-title":"Identifying the best machine learning algorithms for brain tumor segmentation, progression assessment, and overall survival prediction in the brats challenge","author":"Bakas","year":""},{"key":"B6","article-title":"A large annotated medical image dataset for the development and evaluation of segmentation algorithms","author":"Simpson","year":""},{"key":"B7","doi-asserted-by":"publisher","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":"B8","first-page":"234","article-title":"U-net: Convolutional networks for biomedical image segmentation","volume-title":"Medical Image Computing and Computer-Assisted Intervention\u2013MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III 18","author":"Ronneberger","year":"2015"},{"key":"B9","article-title":"Stand-alone self-attention in vision models","volume-title":"Advances in Neural Information Processing Systems","author":"Ramachandran","year":"2019"},{"key":"B10","article-title":"A study on u-net limitations in object localization and image segmentation","volume-title":"Society for Imaging Informatics in Medicine (SIIM), Virtual Meeting","author":"Shahedi","year":"2020"},{"key":"B11","article-title":"Rethinking atrous convolution for semantic image segmentation","author":"Chen","year":""},{"key":"B12","doi-asserted-by":"publisher","first-page":"2281","DOI":"10.1109\/TMI.2019.2903562","article-title":"Ce-net: Context encoder network for 2D medical image segmentation","volume":"38","author":"Gu","year":"2019","journal-title":"IEEE Trans Med Imaging"},{"key":"B13","article-title":"An image is worth 16\u00d716 words: transformers for image recognition at scale","author":"Dosovitskiy","year":""},{"key":"B14","first-page":"213","article-title":"End-to-end object detection with transformers","volume-title":"European Conference on Computer Vision","author":"Carion","year":"2020"},{"key":"B15","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1155\/2017\/6014967","article-title":"Primary and secondary spontaneous pneumothorax: prevalence, clinical features, and in-hospital mortality","volume":"2017","author":"Onuki","year":"2017","journal-title":"Can Respir J"},{"key":"B16","doi-asserted-by":"publisher","DOI":"10.5772\/intechopen.83458","article-title":"Primary spontaneous pneumothorax, a clinical challenge","author":"Giraldo Vallejo","year":"2019","journal-title":"Pneumothorax"},{"key":"B17","doi-asserted-by":"publisher","first-page":"S435","DOI":"10.3978\/j.issn.2072-1439.2014.08.34","article-title":"Acute respiratory distress syndrome and pneumothorax","volume":"6","author":"Terzi","year":"2014","journal-title":"J Thorac Dis"},{"key":"B18","doi-asserted-by":"publisher","first-page":"197","DOI":"10.1186\/1749-8090-8-197","article-title":"Tension pneumothorax, is it a really life-threatening condition?","volume":"8","author":"Yoon","year":"2013","journal-title":"J Cardiothorac Surg"},{"key":"B19","article-title":"Data from: SunStar","year":""},{"key":"B20","doi-asserted-by":"publisher","first-page":"859","DOI":"10.1378\/chest.10-2946","article-title":"Diagnosis of pneumothorax by radiography and ultrasonography: a meta-analysis","volume":"140","author":"Ding","year":"2011","journal-title":"Chest"},{"key":"B21","doi-asserted-by":"publisher","first-page":"265","DOI":"10.4103\/1658-354X.174906","article-title":"Lung ultrasound versus chest radiography for the diagnosis of pneumothorax in critically ill patients: a prospective, single-blind study","volume":"10","author":"Abdalla","year":"2016","journal-title":"Saudi J Anaesth"},{"key":"B22","first-page":"29","article-title":"Diagnostic accuracy of chest ultrasonography versus chest radiography for identification of pneumothorax: a systematic review and meta-analysis","volume":"13","author":"Ebrahimi","year":"2014","journal-title":"Tanaffos"},{"key":"B23","doi-asserted-by":"publisher","first-page":"703","DOI":"10.1378\/chest.11-0131","article-title":"Test characteristics of ultrasonography for the detection of pneumothorax: a systematic review and meta-analysis","volume":"141","author":"Alrajhi","year":"2012","journal-title":"Chest"},{"key":"B24","article-title":"Data from: Siim-acr pneumothorax segmentation","author":"Zawacki","year":"2019"},{"key":"B25","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1007\/978-3-030-00889-5_1","article-title":"Unet++: A nested u-net architecture for medical image segmentation","volume-title":"Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support: 4th International Workshop, DLMIA 2018, and 8th International Workshop, ML-CDS 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 20, 2018, Proceedings 4","author":"Zhou","year":"2018"},{"key":"B26","first-page":"770","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"B27","first-page":"7132","article-title":"Squeeze-and-excitation networks","author":"Hu","year":"2018"},{"key":"B28","first-page":"1492","article-title":"Aggregated residual transformations for deep neural networks","author":"Xie","year":"2017"},{"key":"B29","first-page":"597","article-title":"On the computational complexity of self-attention","volume-title":"International Conference on Algorithmic Learning Theory","author":"Keles","year":"2023"},{"key":"B30","first-page":"10012","article-title":"Swin transformer: hierarchical vision transformer using shifted windows","author":"Liu","year":"2021"},{"key":"B31","first-page":"568","article-title":"Pyramid vision transformer: a versatile backbone for dense prediction without convolutions","author":"Wang","year":"2021"},{"key":"B32","doi-asserted-by":"publisher","first-page":"9355","DOI":"10.5555\/3540261.3540977","article-title":"Twins: revisiting the design of spatial attention in vision transformers","volume":"34","author":"Chu","year":"2021","journal-title":"Adv Neural Inf Process Syst"},{"key":"B33","first-page":"11976","article-title":"A convnet for the 2020s","author":"Liu","year":"2022"},{"key":"B34","article-title":"Gaussian error linear units (gelus)","author":"Hendrycks","year":""},{"key":"B35","first-page":"448","article-title":"Batch normalization: accelerating deep network training by reducing internal covariate shift","volume-title":"International Conference on Machine Learning","author":"Ioffe","year":"2015"},{"key":"B36","article-title":"Layer normalization","author":"Ba","year":""},{"key":"B37","article-title":"Pneumothorax segmentation: deep learning image segmentation to predict pneumothorax","author":"Jakhar","year":""},{"key":"B38","doi-asserted-by":"crossref","first-page":"698","DOI":"10.1109\/CSCI49370.2019.00131","article-title":"Content-adaptive u-net architecture for medical image segmentation","volume-title":"2019 International Conference on Computational Science and Computational Intelligence (CSCI)","author":"Mostayed","year":"2019"},{"key":"B39","first-page":"11166","article-title":"Pixel-adaptive convolutional neural networks","author":"Su","year":"2019"},{"key":"B40","doi-asserted-by":"publisher","first-page":"e0242013","DOI":"10.1371\/journal.pone.0242013","article-title":"Chexlocnet: automatic localization of pneumothorax in chest radiographs using deep convolutional neural networks","volume":"15","author":"Wang","year":"2020","journal-title":"PLoS ONE"},{"key":"B41","first-page":"2961","article-title":"Mask r-cnn","author":"He","year":"2017"},{"key":"B42","first-page":"2117","article-title":"Feature pyramid networks for object detection","author":"Lin","year":"2017"},{"key":"B43","doi-asserted-by":"publisher","first-page":"e607","DOI":"10.7717\/peerj-cs.607","article-title":"Chest x-ray pneumothorax segmentation using u-net with efficientnet and resnet architectures","volume":"7","author":"Abedalla","year":"2021","journal-title":"PeerJ Comput Sci"},{"key":"B44","doi-asserted-by":"publisher","first-page":"2278","DOI":"10.3390\/s22062278","article-title":"Deep learning-based computer-aided pneumothorax detection using chest x-ray images","volume":"22","author":"Malhotra","year":"2022","journal-title":"Sensors"},{"key":"B45","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-031-73661-2_5","article-title":"Mobilenetv4-universal models for the mobile ecosystem","author":"Qin","year":""},{"key":"B46","article-title":"Rectified linear units improve restricted boltzmann machines","volume-title":"ICML","author":"Nair","year":"2010"},{"key":"B47","doi-asserted-by":"publisher","first-page":"3493","DOI":"10.1118\/1.3611983","article-title":"Su-e-t-33: pydicom: an open source dicom library","volume":"38","author":"Mason","year":"2011","journal-title":"Med Phys"},{"key":"B48","article-title":"Data from: Amazoncom I. Amazon sagemaker ground truth (2024)","year":""},{"key":"B49","doi-asserted-by":"crossref","first-page":"248","DOI":"10.1109\/CVPR.2009.5206848","article-title":"Imagenet: a large-scale hierarchical image database","volume-title":"2009 IEEE Conference on Computer Vision and Pattern Recognition","author":"Deng","year":"2009"},{"key":"B50","article-title":"Data from: Segmentation models pytorch. GitHub repository (2020)","author":"Yakubovskiy","year":""},{"key":"B51","first-page":"379","article-title":"Tversky loss function for image segmentation using 3D fully convolutional deep networks","volume-title":"International Workshop on Machine Learning in Medical Imaging","author":"Salehi","year":"2017"},{"key":"B52","first-page":"2980","article-title":"Focal loss for dense object detection","author":"Lin","year":"2017"},{"key":"B53","article-title":"Adam: a method for stochastic optimization","author":"Kingma","year":""},{"key":"B54","article-title":"Data from: Databricks I. Mlflow (2024)","year":""},{"key":"B55","doi-asserted-by":"crossref","first-page":"559","DOI":"10.1007\/978-1-0716-1418-1","volume-title":"An Introduction to Statistical Learning: With Applications in R","author":"James","year":"2021"},{"key":"B56","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1177\/0885066620987813","article-title":"Current trends in the use of ultrasound over chest x-ray to identify pneumothoraces in ICU, trauma, and ards patients","volume":"37","author":"Ron","year":"2022","journal-title":"J Intensive Care Med"}],"container-title":["Frontiers in Radiology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fradi.2024.1424065\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,12,11]],"date-time":"2024-12-11T04:34:27Z","timestamp":1733891667000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fradi.2024.1424065\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,12,11]]},"references-count":56,"alternative-id":["10.3389\/fradi.2024.1424065"],"URL":"https:\/\/doi.org\/10.3389\/fradi.2024.1424065","relation":{},"ISSN":["2673-8740"],"issn-type":[{"value":"2673-8740","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,12,11]]},"article-number":"1424065"}}