Investigation of Appropriate Scaling of Networks and Images for Convolutional Neural Network-Based Nerve Detection in Ultrasound-Guided Nerve Blocks
Sensors · 2024
DOI: 10.3390/s24113696 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, 2024. Verfasst von Takaaki Sugino und 9 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biomedical Informatics, Institute of Biomaterials and Bioengineering, Tokyo Medical and Dental University, Tokyo 101-0062, Japan. Hauptthemen laut MEDLINE: Nerve Block, Neural Networks, Computer, Ultrasonography. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Japan Society for the Promotion of Science KAKENHI, The Cooperative Research Project of Research Center for Biomedical Engineering, "Subsidy for operating cost to promote practical development for area rehabilitation" by METI and Fukushima Prefecture.
Abstract (Originalsprache)
Ultrasound imaging is an essential tool in anesthesiology, particularly for ultrasound-guided peripheral nerve blocks (US-PNBs). However, challenges such as speckle noise, acoustic shadows, and variability in nerve appearance complicate the accurate localization of nerve tissues. To address this issue, this study introduces a deep convolutional neural network (DCNN), specifically Scaled-YOLOv4, and investigates an appropriate network model and input image scaling for nerve detection on ultrasound images. Utilizing two datasets, a public dataset and an original dataset, we evaluated the effects of model scale and input image size on detection performance. Our findings reveal that smaller input images and larger model scales significantly improve detection accuracy. The optimal configuration of model size and input image size not only achieved high detection accuracy but also demonstrated real-time processing capabilities.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Biomedical Informatics, Institute of Biomaterials and Bioengineering, Tokyo Medical and Dental University, Tokyo 101-0062, Japan.
- Department of Anesthesiology, Fukushima Medical University, Fukushima 960-1295, Japan.
- TCC Media Lab Co., Ltd., Tokyo 192-0152, Japan.
- IOT SOFT Co., Ltd., Tokyo 103-0023, Japan.
ORCID-Profile: Takaaki Sugino, Shinya Onogi, Satoki Inoue, Yoshikazu Nakajima
Förderung
- Japan Society for the Promotion of Science KAKENHI
- The Cooperative Research Project of Research Center for Biomedical Engineering
- "Subsidy for operating cost to promote practical development for area rehabilitation" by METI and Fukushima Prefecture
- JSPS KAKENHI
- Subsidy for operating cost to promote practical development for area rehabilitation
- METI and Fukushima Prefecture
- Cooperative Research Project from the Research Center for Biomedical Engineering
Zitieren
APA
Sugino, T., Onogi, S., Oishi, R., Hanayama, C., Inoue, S., Ishida, S., Yao, Y., Ogasawara, N., Murakawa, M., & Nakajima, Y. (2024). Investigation of Appropriate Scaling of Networks and Images for Convolutional Neural Network-Based Nerve Detection in Ultrasound-Guided Nerve Blocks. Sensors. https://doi.org/10.3390/s24113696
Vancouver
Sugino T, Onogi S, Oishi R, Hanayama C, Inoue S, Ishida S, et al. Investigation of Appropriate Scaling of Networks and Images for Convolutional Neural Network-Based Nerve Detection in Ultrasound-Guided Nerve Blocks. Sensors. 2024. doi:10.3390/s24113696
BibTeX
@article{Sugino2024,
author = {Sugino, Takaaki and Onogi, Shinya and Oishi, Rieko and Hanayama, Chie and Inoue, Satoki and Ishida, Shinjiro and Yao, Yuhang and Ogasawara, Nobuhiro and Murakawa, Masahiro and Nakajima, Yoshikazu},
title = {Investigation of Appropriate Scaling of Networks and Images for Convolutional Neural Network-Based Nerve Detection in Ultrasound-Guided Nerve Blocks},
journal = {Sensors},
year = {2024},
doi = {10.3390/s24113696}
}
EUVIMED