EUVIMEDEUVIMED entdecken
Open Access

Detecting Uniformity Artifacts in Ultrasound Transducers

Chunming Gu iD, Kevin Brom, Scott Stekel, Donald J Tradup, Zhuoyao Xin, Nicholas J Hangiandreou, Jaydev K Dave, Zaiyang Long

Journal of Ultrasound in Medicine · 2026-09 · 45 (9) · 1889-1896

DOI: 10.1002/jum.70239 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Ultrasound in Medicine, online erschienen am 01.04.2026. Verfasst von Chunming Gu und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiology, Mayo Clinic, Rochester, Minnesota, USA. Hauptthemen laut MEDLINE: Deep Learning, Artifacts, Transducers, Image Processing, Computer-Assisted. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 21 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objectives

Uniformity artifacts caused by defective transducer elements or scanner malfunctions degrade diagnostic image quality. Traditional quality control (QC) methods, such as phantom testing and visual or quantitative image analysis, can be labor‐intensive and limited in test frequency. This study aims to develop a deep learning framework to detect uniformity artifacts and complement traditional QC.

Methods

Clinical median images were generated by aggregating co‐registered grayscale ultrasound images acquired by each transducer and computing median intensity values across the image stack. A pretrained ResNet‐18 model was fine‐tuned on a dataset consisting of clinical median images from linear and curvilinear transducers. The dataset was divided into training, validation, and testing subsets, ensuring no overlap between training and test transducers. To assess generalizability, the model was also evaluated on an independent test set of 396 phantom‐validated images from linear and curvilinear transducers.

Results

The model achieved 100% accuracy on the first dataset's test set. On the independent test set, it attained 87.4% accuracy with high sensitivity (0.84) and specificity (0.88), demonstrating robust generalization. Occlusion sensitivity maps confirmed the model's attention to uniformity artifact regions.

Conclusion

The deep learning framework using clinical median images demonstrated robust performance across several linear and curvilinear transducer models. It could be integrated into the clinical QC workflow by automating artifact detection in an effective and timely manner. In our practice, it can flag median images classified as artifact‐present, reducing human review time by approximately 80% while preserving detection accuracy.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningArtifactsTransducersImage Processing, Computer-AssistedUltrasonographyHumansPhantoms, ImagingSensitivity and SpecificityReproducibility of ResultsDetection Algorithms

Stichwörter der Autor:innen

automated artifact detectiondeep learningimage processingultrasound Imaginguniformity artifacts

Autor:innen und Institutionen

  • Department of Radiology, Mayo Clinic, Rochester, Minnesota, USA.
  • Department of Biomedical Engineering, Johns Hopkins University, Baltimore, Maryland, USA.

ORCID-Profile: Chunming Gu

Zitieren

APA

Gu, C., Brom, K., Stekel, S., Tradup, D. J., Xin, Z., Hangiandreou, N. J., Dave, J. K., & Long, Z. (2026). Detecting Uniformity Artifacts in Ultrasound Transducers. Journal of Ultrasound in Medicine, 45(9), 1889-1896. https://doi.org/10.1002/jum.70239

Vancouver

Gu C, Brom K, Stekel S, Tradup DJ, Xin Z, Hangiandreou NJ, et al. Detecting Uniformity Artifacts in Ultrasound Transducers. Journal of Ultrasound in Medicine. 2026;45(9):1889-1896. doi:10.1002/jum.70239

BibTeX
@article{Gu2026,
  author = {Gu, Chunming and Brom, Kevin and Stekel, Scott and Tradup, Donald J and Xin, Zhuoyao and Hangiandreou, Nicholas J and Dave, Jaydev K and Long, Zaiyang},
  title = {Detecting Uniformity Artifacts in Ultrasound Transducers},
  journal = {Journal of Ultrasound in Medicine},
  year = {2026},
  volume = {45},
  number = {9},
  pages = {1889--1896},
  doi = {10.1002/jum.70239}
}
Kostenlos starten →