EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Deep learning cascade networks for segmentation of fluorine-18 sodium fluoride positron emission tomography scans of equine metacarpo- and metatarsophalangeal joints outperform atlas-based method

Sergey Anishchenko iD, Kathryn W Bills, David Beylin, Natalie Beylin, Danielle Beylin, Karina Stepanova, Pavel Stepanov, Mathieu Spriet iD

American Journal of Veterinary Research · 2026

DOI: 10.2460/ajvr.25.11.0421 ↗

Kurz gefasst

Fachartikel in der Zeitschrift American Journal of Veterinary Research, 2026. Verfasst von Sergey Anishchenko und 7 weiteren Autor:innen. Beteiligte Einrichtung: Alienbyte Scientific Software Inc, Rockville, MD. Hauptthemen laut MEDLINE: Positron-Emission Tomography, Deep Learning, Metatarsophalangeal Joint, Metacarpophalangeal Joint. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objective

To create a labeled dataset and evaluate a convolutional neural network (CNN) for segmentation of fluorine-18 sodium fluoride PET scans of the equine metacarpo- and metatarsophalangeal joint (fetlock), targeting the third metacarpal bone, proximal phalanx, proximal sesamoid bone(s), and soft tissue.

Methods

PET and corresponding CT scans were retrospectively selected from June 2024 through November 2025 using convenience sampling and coregistered to a common reference frame. The coregistered PET-CT scans were manually labeled; the labels were then transferred to PET-only images and used to train a cascade of CNNs with and without data augmentation. Segmentation accuracy was quantified using the Dice coefficient and compared to atlas-based segmentation of the PET-CT scans.

Results

A total of 84 scans were collected from 2 facilities. For the test set (n = 8), the minimum, mean, and maximum Dice coefficients computed for all anatomical areas together were 0.85, 0.88, and 0.92, respectively, exceeding the values for atlas-based segmentation of 0.69, 0.80, and 0.87. Analysis of the Dice coefficient by area showed the best CNN segmentation for the third metacarpal bone (mean Dice coefficient (D mean ) = 0.92) and the worst for the medial proximal sesamoid bone (D mean = 0.82); corresponding atlas-based values were 0.86 and 0.77.

Conclusions

CNN-based segmentation of fluorine-18 sodium fluoride PET images of the fetlock outperforms an atlas-based method in both speed and accuracy measured by Dice coefficient. This advancement enables new, previously inaccessible strategies for PET image quantification. Clinical Relevance Accurate and robust segmentation of PET fetlocks enables more accurate analysis and novel insights into lesion characterization.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Positron-Emission TomographyDeep LearningMetatarsophalangeal JointMetacarpophalangeal JointAnimalsHorsesSodium FluorideFluorine RadioisotopesRetrospective StudiesConvolutional Neural Networks

Stichwörter der Autor:innen

AICNNequine positron emission tomographymedical imagesegmentation

Substanzen

Sodium FluorideFluorine RadioisotopesFluorine-18

Autor:innen und Institutionen

  • Alienbyte Scientific Software Inc, Rockville, MD.
  • LONGMILE Veterinary Imaging, Rockville, MD.
  • Department of Clinical Studies, New Bolton Center, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA.
  • Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California-Davis, Davis, CA.

ORCID-Profile: Sergey Anishchenko, Mathieu Spriet

Zitieren

APA

Anishchenko, S., Bills, K. W., Beylin, D., Beylin, N., Beylin, D., Stepanova, K., Stepanov, P., & Spriet, M. (2026). Deep learning cascade networks for segmentation of fluorine-18 sodium fluoride positron emission tomography scans of equine metacarpo- and metatarsophalangeal joints outperform atlas-based method. American Journal of Veterinary Research. https://doi.org/10.2460/ajvr.25.11.0421

Vancouver

Anishchenko S, Bills KW, Beylin D, Beylin N, Beylin D, Stepanova K, et al. Deep learning cascade networks for segmentation of fluorine-18 sodium fluoride positron emission tomography scans of equine metacarpo- and metatarsophalangeal joints outperform atlas-based method. American Journal of Veterinary Research. 2026. doi:10.2460/ajvr.25.11.0421

BibTeX
@article{Anishchenko2026,
  author = {Anishchenko, Sergey and Bills, Kathryn W and Beylin, David and Beylin, Natalie and Beylin, Danielle and Stepanova, Karina and Stepanov, Pavel and Spriet, Mathieu},
  title = {Deep learning cascade networks for segmentation of fluorine-18 sodium fluoride positron emission tomography scans of equine metacarpo- and metatarsophalangeal joints outperform atlas-based method},
  journal = {American Journal of Veterinary Research},
  year = {2026},
  doi = {10.2460/ajvr.25.11.0421}
}
Kostenlos starten →