{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T02:02:09Z","timestamp":1785463329420,"version":"3.56.0"},"reference-count":36,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,4,22]],"date-time":"2025-04-22T00:00:00Z","timestamp":1745280000000},"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. Artif. Intell."],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>Lung ultrasound (LUS) has become an essential imaging modality for assessing various pulmonary conditions, including the presence of B-line artifacts. These artifacts are commonly associated with conditions such as increased extravascular lung water, decompensated heart failure, dialysis-related chronic kidney disease, interstitial lung disease, and COVID-19 pneumonia. Accurate detection of the B-line in LUS images is crucial for effective diagnosis and treatment. However, interpreting LUS is often subject to observer variability, requiring significant expertise and posing challenges in resource-limited settings with few trained professionals.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>To address these limitations, deep learning models have been developed for automated B-line detection and localization. This study introduces YOLOv5-PBB and YOLOv8-PBB, two modified models based on YOLOv5 and YOLOv8, respectively, designed for precise and interpretable B-line localization using polygonal bounding boxes (PBBs). YOLOv5-PBB was enhanced by modifying the detection head, loss function, non-maximum suppression, and data loader to enable PBB localization. YOLOv8-PBB was customized to convert segmentation masks into polygonal representations, displaying only boundaries while removing the masks. Additionally, an image preprocessing technique was incorporated into the models to enhance LUS image quality. The models were trained on a diverse dataset from a publicly available repository and Ugandan health facilities.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Experimental results showed that YOLOv8-PBB achieved the highest precision (0.947), recall (0.926), and mean average precision (0.957). YOLOv5-PBB, while slightly lower in performance (precision: 0.931, recall: 0.918, mAP: 0.936), had advantages in model size (14 MB vs. 21 MB) and average inference time (33.1 ms vs. 47.7 ms), making it more suitable for real-time applications in low-resource settings.<\/jats:p><\/jats:sec><jats:sec><jats:title>Discussion<\/jats:title><jats:p>The integration of these models into a mobile LUS screening tool provides a promising solution for B-line localization in resource-limited settings, where accessibility to trained professionals may be scarce. The YOLOv5-PBB and YOLOv8-PBB models offer high performance while addressing challenges related to inference speed and model size, making them ideal candidates for mobile deployment in such environments.<\/jats:p><\/jats:sec>","DOI":"10.3389\/frai.2025.1560523","type":"journal-article","created":{"date-parts":[[2025,4,22]],"date-time":"2025-04-22T05:27:33Z","timestamp":1745299653000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["Deep learning for accurate B-line detection and localization in lung ultrasound imaging"],"prefix":"10.3389","volume":"8","author":[{"given":"Nixson","family":"Okila","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andrew","family":"Katumba","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joyce","family":"Nakatumba-Nabende","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cosmas","family":"Mwikirize","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sudi","family":"Murindanyi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jonathan","family":"Serugunda","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Samuel","family":"Bugeza","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Anthony","family":"Oriekot","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Juliet","family":"Bossa","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Eva","family":"Nabawanuka","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,4,22]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"318","DOI":"10.1148\/radiol.2018171820","article-title":"Current applications and future impact of machine learning in radiology","volume":"288","author":"Choy","year":"2018","journal-title":"Radiology"},{"key":"B2","doi-asserted-by":"publisher","first-page":"2312","DOI":"10.1109\/TUFFC.2020.3002249","article-title":"Automated lung ultrasound B-line assessment using a deep learning algorithm","volume":"67","author":"Cristiana","year":"2020","journal-title":"IEEE Trans. Ultrason. Ferroelectr. Frequency Control"},{"key":"B3","doi-asserted-by":"publisher","first-page":"309","DOI":"10.1002\/jum.16088","article-title":"New international guidelines and consensus on the use of lung ultrasound","volume":"42","author":"Demi","year":"2023","journal-title":"J. Ultrasound Med"},{"key":"B4","doi-asserted-by":"publisher","first-page":"612","DOI":"10.1109\/TUFFC.2019.2949597","article-title":"Physical mechanisms providing clinical information from ultrasound lung images: hypotheses and early confirmations","volume":"67","author":"Demi","year":"2020","journal-title":"IEEE Trans. Ultrason. Ferroelectr. Frequency Control"},{"key":"B5","doi-asserted-by":"publisher","first-page":"515","DOI":"10.1016\/j.jemermed.2020.06.033","article-title":"Best practice recommendations for point-of-care lung ultrasound in patients with suspected COVID-19","volume":"59","author":"Duggan","year":"2020","journal-title":"J. Emerg. Med"},{"key":"B6","doi-asserted-by":"crossref","first-page":"2276","DOI":"10.1145\/3343031.3350535","article-title":"\u201cThe via annotation software for images, audio and video,\u201d","volume-title":"Proceedings of the 27th ACM International Conference on Multimedia","author":"Dutta","year":"2019"},{"key":"B7","doi-asserted-by":"publisher","first-page":"2097","DOI":"10.1002\/jum.15894","article-title":"Lung ultrasound for pleural line abnormalities, confluent b-lines, and consolidation","volume":"41","author":"Fischer","year":"2022","journal-title":"J. Ultrasound Med"},{"key":"B8","doi-asserted-by":"publisher","first-page":"2483","DOI":"10.1002\/jum.15338","article-title":"Contrast\u2013enhanced ultrasound in patients with COVID\u201319","volume":"49","author":"Gino","year":"2020","journal-title":"J. Ultrasound Med"},{"key":"B9","volume-title":"Automatic Detection of B-lines in Lung Ultrasound Videos from Severe Dengue Patients","author":"Hamideh","year":"2021"},{"key":"B10","doi-asserted-by":"publisher","first-page":"34","DOI":"10.1097\/AIA.0000000000000366","article-title":"An introduction to lung ultrasound","volume":"60","author":"Hernandez-Morgan","year":"2022","journal-title":"Int. Anesthesiol. Clinics"},{"key":"B11","doi-asserted-by":"publisher","first-page":"107251","DOI":"10.1016\/j.ultras.2024.107251","article-title":"Deep learning for real-time multi-class segmentation of artefacts in lung ultrasound","volume":"140","author":"Howell","year":"2024","journal-title":"Ultrasonics"},{"key":"B12","doi-asserted-by":"publisher","first-page":"2585","DOI":"10.1007\/s10115-021-01605-0","article-title":"Model complexity of deep learning: a survey","volume":"63","author":"Hu","year":"2021","journal-title":"Knowl. Inf. Syst"},{"key":"B13","doi-asserted-by":"publisher","first-page":"42816","DOI":"10.1109\/ACCESS.2024.3378568","article-title":"Yolov1 to v8: unveiling each variant-a comprehensive review of yolo","volume":"12","author":"Hussain","year":"2024","journal-title":"IEEE Access"},{"key":"B14","author":"Jocher","year":"2023","journal-title":"Yolo by Ultralytics (Version 8.0. 0)[Computer Software"},{"key":"B15","author":"Jocher","year":"2021","journal-title":"Ultralytics\/yolov5: v5. 0-yolov5-p6 1280 Models, Aws, Supervise. Ly and Youtube Integrations"},{"key":"B16","doi-asserted-by":"publisher","first-page":"100094","DOI":"10.1016\/j.bea.2023.100094","article-title":"Lungecho-resource constrained lung ultrasound video analysis tool for faster triaging and active learning","volume":"6","author":"Joseph","year":"2023","journal-title":"Biomed. Eng. Adv"},{"key":"B17","first-page":"6517","article-title":"\u201cYou only look once: unified, real-time object detection,\u201d","volume-title":"2017 IEEE Conference on Computer Vision and Pattern Recognition","author":"Joseph","year":"2017"},{"key":"B18","doi-asserted-by":"publisher","first-page":"176346","DOI":"10.1109\/ACCESS.2024.3505619","article-title":"Real-time polyp detection from endoscopic images using yolov8 with yolo-score metrics for enhanced suitability assessment","volume":"12","author":"Khan","year":"2024","journal-title":"IEEE Access"},{"key":"B19","doi-asserted-by":"publisher","first-page":"71739","DOI":"10.1109\/ACCESS.2020.2987868","article-title":"Fisheyedet: a self-study and contour-based object detector in fisheye images","volume":"1","author":"Li","year":"2020","journal-title":"IEEE Access"},{"key":"B20","doi-asserted-by":"publisher","first-page":"4352","DOI":"10.1109\/JBHI.2023.3282596","article-title":"Deep learning for detection and localization of b-lines in lung ultrasound","volume":"27","author":"Lucassen","year":"2023","journal-title":"IEEE J. Biomed. Health Informat"},{"key":"B21","doi-asserted-by":"publisher","first-page":"1515","DOI":"10.1093\/rheumatology\/keab801","article-title":"Lung ultrasound b-lines in systemic sclerosis: cut-off values and methodological indications for interstitial lung disease screening","volume":"158","author":"Luna","year":"2022","journal-title":"Rheumatology"},{"key":"B22","doi-asserted-by":"publisher","first-page":"e200564","DOI":"10.1148\/ryct.2021200564","article-title":"Lung ultrasound: the essentials","volume":"3","author":"Marini","year":"2021","journal-title":"Radiol. Cardiothorac. Imaging"},{"key":"B23","doi-asserted-by":"publisher","first-page":"1","DOI":"10.22514\/sv.2020.16.0041","article-title":"The role of b-line artifacts on lung ultrasound in critically ill patients","volume":"16","author":"Piculjan","year":"2020","journal-title":"Signa Vitae"},{"key":"B24","first-page":"3103","article-title":"\u201cDeep learning video classification of lung ultrasound features associated with pneumonia,\u201d","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Shea","year":"2023"},{"key":"B25","doi-asserted-by":"publisher","first-page":"163","DOI":"10.1080\/17476348.2019.1565997","article-title":"The role of ultrasound lung artifacts in the diagnosis of respiratory diseases","volume":"13","author":"Soldati","year":"2019","journal-title":"Expert Rev. Respir. Med"},{"key":"B26","doi-asserted-by":"publisher","first-page":"1459","DOI":"10.1002\/jum.15284","article-title":"Is there a role for lung ultrasound during the covid-19 pandemic?","volume":"2022","author":"Soldati","year":"2020","journal-title":"J. Ultrasound Med"},{"key":"B27","first-page":"65","article-title":"\u201cUltrasound-based detection of lung abnormalities using single shot detection convolutional neural networks,\u201d","volume-title":"Proceedings of the MICCAI-PoCUS","author":"Sourabh","year":"2018"},{"key":"B28","doi-asserted-by":"publisher","first-page":"2676","DOI":"10.1109\/TMI.2020.2994459","article-title":"Deep learning for classification and localization of covid-19 markers in point-of-care lung ultrasound","volume":"39","author":"Subhankar","year":"2020","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B29","doi-asserted-by":"publisher","first-page":"3668","DOI":"10.3390\/cancers16213668","article-title":"Grad-cam enabled breast cancer classification with a 3d inception-resnet v2: empowering radiologists with explainable insights","volume":"16","author":"Talaat","year":"2024","journal-title":"Cancers"},{"key":"B30","doi-asserted-by":"publisher","first-page":"4793","DOI":"10.1109\/TNNLS.2020.3027314","article-title":"A survey on explainable artificial intelligence (Xai): toward medical Xai","volume":"32","author":"Tjoa","year":"2020","journal-title":"IEEE Trans. Neural Networks Learn. Syst"},{"key":"B31","doi-asserted-by":"publisher","first-page":"957","DOI":"10.1109\/JBHI.2019.2936151","article-title":"Localizing B-lines in lung ultrasonography by weakly-supervised deep learning","volume":"24","author":"Van Sloun","year":"2020","journal-title":"IEEE J. Biomed. Health Informat"},{"key":"B32","doi-asserted-by":"publisher","first-page":"2","DOI":"10.1109\/TUFFC.2022.3221682","article-title":"Current advances in computational lung ultrasound imaging: a review","volume":"70","author":"Yang","year":"2023","journal-title":"IEEE Trans. Ultrason. Ferroelectr. Frequency Control"},{"key":"B33","doi-asserted-by":"publisher","first-page":"18381","DOI":"10.48550\/arXiv.2106.01883","article-title":"Learning high-precision bounding box for rotated object detection via kullback-leibler divergence","volume":"34","author":"Yang","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst"},{"key":"B34","doi-asserted-by":"publisher","first-page":"3886","DOI":"10.1109\/TAI.2024.3353157","article-title":"Complexity-driven model compression for resource-constrained deep learning on edge","volume":"5","author":"Zawish","year":"2024","journal-title":"IEEE Trans. Artif. Intell"},{"key":"B35","first-page":"1","article-title":"\u201cCOVID-19 feature detection with deep neural networks trained on simulated lung ultrasound b-mode images,\u201d","volume-title":"2022 IEEE International Ultrasonics Symposium (IUS)","author":"Zhao","year":"2022"},{"key":"B36","doi-asserted-by":"publisher","first-page":"187","DOI":"10.1007\/s42058-024-00158-9","article-title":"Advances in multimodal imaging of lung injury","volume":"7","author":"Zhou","year":"2024","journal-title":"Chin. J. Acad. Radiol"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1560523\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,4,22]],"date-time":"2025-04-22T05:27:36Z","timestamp":1745299656000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1560523\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4,22]]},"references-count":36,"alternative-id":["10.3389\/frai.2025.1560523"],"URL":"https:\/\/doi.org\/10.3389\/frai.2025.1560523","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4,22]]},"article-number":"1560523"}}