Deep Learning for FAST Quality Assessment.
FAST
autoencoder
convolutional neural network
deep learning
ultrasound
Journal
Journal of ultrasound in medicine : official journal of the American Institute of Ultrasound in Medicine
ISSN: 1550-9613
Titre abrégé: J Ultrasound Med
Pays: England
ID NLM: 8211547
Informations de publication
Date de publication:
Jan 2023
Jan 2023
Historique:
revised:
30
04
2022
received:
13
10
2021
accepted:
04
06
2022
pubmed:
1
7
2022
medline:
22
12
2022
entrez:
30
6
2022
Statut:
ppublish
Résumé
To determine the feasibility of using a deep learning (DL) algorithm to assess the quality of focused assessment with sonography in trauma (FAST) exams. Our dataset consists of 441 FAST exams, classified as good-quality or poor-quality, with 3161 videos. We first used convolutional neural networks (CNNs), pretrained on the Imagenet dataset and fine-tuned on the FAST dataset. Second, we trained a CNN autoencoder to compress FAST images, with a 20-1 compression ratio. The compressed codes were input to a two-layer classifier network. To train the networks, each video was labeled with the quality of the exam, and the frames were labeled with the quality of the video. For inference, a video was classified as poor-quality if half the frames were classified as poor-quality by the network, and an exam was classified as poor-quality if half the videos were classified as poor-quality. The results with the encoder-classifier networks were much better than the transfer learning results with CNNs. This was primarily because the Imagenet dataset is not a good match for the ultrasound quality assessment problem. The DL models produced video sensitivities and specificities of 99% and 98% on held-out test sets. Using an autoencoder to compress FAST images is a very effective way to obtain features that can be used to predict exam quality. These features are more suitable than those obtained from CNNs pretrained on Imagenet.
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
71-79Subventions
Organisme : Prisma Health Transformative Research Seed Grant program
Informations de copyright
© 2022 American Institute of Ultrasound in Medicine.
Références
Bahner D, Blaivas M, Cohen HL, et al. American Institute of Ultrasound in Medicine. AIUM practice guideline for the performance of the focused assessment with sonography for trauma (FAST) examination. J Ultrasound Med 2008; 27:313-318.
Lewiss RE, Tayal VS, Hoffmann B, et al. The core content of clinical ultrasonography fellowship training. Acad Emerg Med 2014; 21:456-461.
Morrow D, Cupp J, Schrift D, Nathanson R, Soni NJ. Point-of-care ultrasound in established settings. South Med J 2018; 111:373-381.
Steinemann S, Fernandez M. Variation in training and use of the focused assessment with sonography in trauma (FAST). Am J Surg 2018; 215:255-258.
POLICY STATEMENT Ultrasound Guidelines: Emergency, point-of-care, and clinical ultrasound guidelines in medicine. Published online 2016.
American College of Emergency Physician/Clinical Ultrasound Accreditation Program. Accessed January 17, 2022. https://webapps.acep.org/cuap/#/home.
Trauma Quality Improvement Program (TQIP). Accessed January 17, 2022. https://www.facs.org/quality-programs/trauma/tqp/center-programs/tqip.
Baugher KM, Euerle BD, Sommerkamp SK, Witting MD. Image quality evaluation of a portable handheld ultrasound machine for the focused assessment with sonography for trauma examination. Am J Emerg Med 2014; 32:389-391.
Ziesmann MT, Park J, Unger BJ, et al. Validation of the quality of ultrasound imaging and competence (QUICk) score as an objective assessment tool for the FAST examination. J Trauma Acute Care Surg 2015; 78:1008-1013.
Shen YT, Chen L, Yue WW, Xu HX. Artificial intelligence in ultrasound. Eur J Radiol 2021; 12:109717.
Brattain LJ, Telfer BA, Dhyani M, Grajo JR, Samir AE. Machine learning for medical ultrasound: status, methods, and future opportunities. Abdomin Radiol 2018; 43:786-799.
Wu L, Cheng JZ, Li S, Lei B, Wang T, Ni D. FUIQA: fetal ultrasound image quality assessment with deep convolutional networks. IEEE Trans Cybern 2017; 47:1336-1349.
Camps SM, Houben T, Carneiro G, et al. Automatic quality assessment of transperineal ultrasound images of the male pelvic region, using deep learning. Ultrasound Med Biol 2020; 46:445-454.
LeCun Y, Bottou L, Bengio Y, Haffner P. Gradient-based learning applied to document recognition. Proc IEEE 1998; 86:2278-2324.
Indolia S, Goswami AK, Mishra SP, Asopa P. Conceptual understanding of convolutional neural network-a deep learning approach. Procedia Comput Sci 2018; 132:679-688.
Byra M, Styczynski G, Szmigielski C, et al. Transfer learning with deep convolutional neural network for liver steatosis assessment in ultrasound images. Int J Comput Assist Radiol Surg 2018; 13:1895-1903.
Deng J, Dong W, Socher R, Li LJ, Li K, Fei-Fei L. Imagenet: a large-scale hierarchical image database. Proc IEEE Conf Comput Vis Pattern Recognit 2009; 20:248-255.
Blaivas M, Blaivas L. Are all deep learning architectures alike for point-of-care ultrasound?: evidence from a cardiac image classification model suggests otherwise. J Ultrasound Med 2020; 39:1187-1194.
Blaivas L, Blaivas M. Are convolutional neural networks trained on ImageNet images wearing rose-colored glasses?: a quantitative comparison of ImageNet, computed tomographic, magnetic resonance, chest X-ray, and point-of-care ultrasound images for quality. J Ultrasound Med 2021; 40:377-383.
Tjoa E, Guan C. A survey on explainable artificial intelligence (xai): toward medical xai. IEEE Trans Neural Netw Learn Syst 2020; 32:4793-4813.