{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T21:52:04Z","timestamp":1792792324473,"version":"4.5.0"},"reference-count":34,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,6,6]],"date-time":"2024-06-06T00:00:00Z","timestamp":1717632000000},"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. Vet. Sci."],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>Military working dogs (MWDs) are essential for military operations in a wide range of missions. With this pivotal role, MWDs can become casualties requiring specialized veterinary care that may not always be available far forward on the battlefield. Some injuries such as pneumothorax, hemothorax, or abdominal hemorrhage can be diagnosed using point of care ultrasound (POCUS) such as the Global FAST\u00ae exam. This presents a unique opportunity for artificial intelligence (AI) to aid in the interpretation of ultrasound images. In this article, deep learning classification neural networks were developed for POCUS assessment in MWDs.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>Images were collected in five MWDs under general anesthesia or deep sedation for all scan points in the Global FAST\u00ae exam. For representative injuries, a cadaver model was used from which positive and negative injury images were captured. A total of 327 ultrasound clips were captured and split across scan points for training three different AI network architectures: MobileNetV2, DarkNet-19, and ShrapML. Gradient class activation mapping (GradCAM) overlays were generated for representative images to better explain AI predictions.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Performance of AI models reached over 82% accuracy for all scan points. The model with the highest performance was trained with the MobileNetV2 network for the cystocolic scan point achieving 99.8% accuracy. Across all trained networks the diaphragmatic hepatorenal scan point had the best overall performance. However, GradCAM overlays showed that the models with highest accuracy, like MobileNetV2, were not always identifying relevant features. Conversely, the GradCAM heatmaps for ShrapML show general agreement with regions most indicative of fluid accumulation.<\/jats:p><\/jats:sec><jats:sec><jats:title>Discussion<\/jats:title><jats:p>Overall, the AI models developed can automate POCUS predictions in MWDs. Preliminarily, ShrapML had the strongest performance and prediction rate paired with accurately tracking fluid accumulation sites, making it the most suitable option for eventual real-time deployment with ultrasound systems. Further integration of this technology with imaging technologies will expand use of POCUS-based triage of MWDs.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fvets.2024.1374890","type":"journal-article","created":{"date-parts":[[2024,6,11]],"date-time":"2024-06-11T19:27:53Z","timestamp":1718134073000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":5,"title":["Deep learning models for interpretation of point of care ultrasound in military working dogs"],"prefix":"10.3389","volume":"11","author":[{"given":"Sofia I.","family":"Hernandez Torres","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lawrence","family":"Holland","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thomas H.","family":"Edwards","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Emilee C.","family":"Venn","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Eric J.","family":"Snider","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2024,6,6]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"773","DOI":"10.1016\/j.cvsm.2013.03.011","article-title":"The use of ultrasound for dogs and cats in the emergency room: AFAST and TFAST","volume":"43","author":"Boysen","year":"2013","journal-title":"Vet Clin North Am Small Anim Pract"},{"key":"ref2","doi-asserted-by":"publisher","first-page":"178","DOI":"10.1111\/jsap.13271","article-title":"Diagnostic accuracy of a lung ultrasound protocol (vet BLUE) for detection of pleural fluid, pneumothorax and lung pathology in dogs and cats","volume":"62","author":"Cole","year":"2021","journal-title":"J Small Anim Pract"},{"key":"ref3","first-page":"52","author":"Boatright","year":"2020"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"1315","DOI":"10.1016\/j.cvsm.2021.07.011","article-title":"Global FAST for patient monitoring and staging in dogs and cats","volume":"51","author":"Lisciandro","year":"2021","journal-title":"Vet Clin North Am Small Anim Pract"},{"key":"ref5","author":"Green","year":"2021"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"498","DOI":"10.1111\/vec.13064","article-title":"Comparison of trauma sustained by civilian dogs and deployed military working dogs","volume":"31","author":"Edwards","year":"2021","journal-title":"J Vet Emerg Crit Care (San Antonio)"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"933","DOI":"10.1016\/j.cvsm.2021.04.010","article-title":"Military working dogs: an overview of veterinary Care of these Formidable Assets","volume":"51","author":"McGraw","year":"2021","journal-title":"Vet Clin North Am Small Anim Pract"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"180","DOI":"10.1093\/milmed\/usy119","article-title":"Trauma Management of Military Working Dogs","volume":"183","author":"Lagutchik","year":"2018","journal-title":"Mil Med"},{"key":"ref9","volume-title":"The U.S. Army in multi-domain operations 2028","author":"Townsend","year":"2018"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"364","DOI":"10.1097\/XCS.0000000000000707","article-title":"Putting medical boots on the ground: lessons from the war in Ukraine and applications for future conflict with near-peer adversaries","volume":"237","author":"Epstein","year":"2023","journal-title":"J Am Coll Surg"},{"key":"ref11","first-page":"1","article-title":"Medical imaging using machine learning and deep learning algorithms: a review","author":"Latif","year":"2019"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"42","DOI":"10.1016\/j.glohj.2020.04.002","article-title":"A review of medical artificial intelligence","volume":"4","author":"Liu","year":"2020","journal-title":"Glob Health J"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"720","DOI":"10.3390\/biomedicines9070720","article-title":"Towards clinical application of artificial intelligence in ultrasound imaging","volume":"9","author":"Komatsu","year":"2021","journal-title":"Biomedicines"},{"key":"ref14","doi-asserted-by":"publisher","first-page":"261","DOI":"10.1016\/j.eng.2018.11.020","article-title":"Deep learning in medical ultrasound analysis: a review","volume":"5","author":"Liu","year":"2019","journal-title":"Engineering"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"240","DOI":"10.1109\/TMI.2018.2860257","article-title":"Tumor detection in automated breast ultrasound using 3-D CNN and prioritized candidate aggregation","volume":"38","author":"Chiang","year":"2019","journal-title":"IEEE Trans Med Imaging"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"107527","DOI":"10.1016\/j.cmpb.2023.107527","article-title":"A new xAI framework with feature explainability for tumors decision-making in ultrasound data: comparing with grad-CAM","volume":"235","author":"Song","year":"2023","journal-title":"Comput Methods Prog Biomed"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"e0255886","DOI":"10.1371\/journal.pone.0255886","article-title":"Deep-learning based detection of COVID-19 using lung ultrasound imagery","volume":"16","author":"Diaz-Escobar","year":"2021","journal-title":"PLoS One"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"110156","DOI":"10.1016\/j.ejrad.2022.110156","article-title":"Ultrasound findings of lung ultrasonography in COVID-19: a systematic review","volume":"148","author":"Gil-Rodr\u00edguez","year":"2022","journal-title":"Eur J Radiol"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"2204","DOI":"10.1109\/TMI.2017.2712367","article-title":"SonoNet: real-time detection and localisation of Fetal standard scan planes in freehand ultrasound","volume":"36","author":"Baumgartner","year":"2017","journal-title":"IEEE Trans Med Imaging"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"101150","DOI":"10.1016\/j.imu.2022.101150","article-title":"Deep learning-based real time detection for cardiac objects with fetal ultrasound video","volume":"36","author":"Iriani Sapitri","year":"2023","journal-title":"Inform Med Unlocked"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"3964","DOI":"10.1038\/s41598-021-83515-3","article-title":"Automatic classification of canine thoracic radiographs using deep learning","volume":"11","author":"Banzato","year":"2021","journal-title":"Sci Rep"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"573","DOI":"10.1111\/vru.13089","article-title":"Accuracy of artificial intelligence software for the detection of confirmed pleural effusion in thoracic radiographs in dogs","volume":"63","author":"M\u00fcller","year":"2022","journal-title":"Vet Radiol Ultrasound"},{"key":"ref23","doi-asserted-by":"publisher","first-page":"292","DOI":"10.1111\/vru.13062","article-title":"Comparison of artificial intelligence to the veterinary radiologist\u2019s diagnosis of canine cardiogenic pulmonary edema","volume":"63","author":"Kim","year":"2022","journal-title":"Vet Radiol Ultrasound"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"611","DOI":"10.1111\/vru.12901","article-title":"Pilot study: application of artificial intelligence for detecting left atrial enlargement on canine thoracic radiographs","volume":"61","author":"Li","year":"2020","journal-title":"Vet Radiol Ultrasound"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"387","DOI":"10.1111\/vru.12968","article-title":"Deep transfer learning can be used for the detection of hip joints in pelvis radiographs and the classification of their hip dysplasia status","volume":"62","author":"McEvoy","year":"2021","journal-title":"Vet Radiol Ultrasound"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"140","DOI":"10.3390\/jimaging8050140","article-title":"Comparison of ultrasound image classifier deep learning algorithms for shrapnel detection","volume":"8","author":"Boice","year":"2022","journal-title":"J Imaging"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"8427","DOI":"10.1038\/s41598-022-12367-2","article-title":"An image classification deep-learning algorithm for shrapnel detection from ultrasound images","volume":"12","author":"Snider","year":"2022","journal-title":"Sci Rep"},{"key":"ref28","doi-asserted-by":"publisher","first-page":"1244616","DOI":"10.3389\/fbioe.2023.1244616","article-title":"An extended focused assessment with sonography in trauma ultrasound tissue-mimicking phantom for developing automated diagnostic technologies","volume":"11","author":"Hernandez-Torres","year":"2023","journal-title":"Front Bioeng Biotechnol"},{"key":"ref29","first-page":"2010.01267","article-title":"Wemix: How to better utilize data augmentation","volume-title":"arXiv preprint","author":"Xu","year":"2020"},{"key":"ref30","first-page":"4510","article-title":"Mobilenetv2: Inverted residuals and linear bottlenecks","volume-title":"Proceedings of the IEEE conference on computer vision and pattern recognition","author":"Sandler","year":"2018"},{"key":"ref31","author":"Redmon","year":"2016"},{"key":"ref32","first-page":"618","article-title":"Grad-cam: visual explanations from deep networks via gradient-based localization","author":"Selvaraju","year":"2017"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"298","DOI":"10.1186\/s12911-022-02047-6","article-title":"Automatic ovarian tumors recognition system based on ensemble convolutional neural network with ultrasound imaging","volume":"22","author":"Hsu","year":"2022","journal-title":"BMC Med Inform Decis Mak"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"905955","DOI":"10.3389\/fonc.2022.905955","article-title":"Ultrasound image classification of thyroid nodules based on deep learning","volume":"12","author":"Yang","year":"2022","journal-title":"Front Oncol"}],"container-title":["Frontiers in Veterinary Science"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fvets.2024.1374890\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,6,11]],"date-time":"2024-06-11T19:27:56Z","timestamp":1718134076000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fvets.2024.1374890\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,6,6]]},"references-count":34,"alternative-id":["10.3389\/fvets.2024.1374890"],"URL":"https:\/\/doi.org\/10.3389\/fvets.2024.1374890","relation":{},"ISSN":["2297-1769"],"issn-type":[{"value":"2297-1769","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,6,6]]},"article-number":"1374890"}}