Validation of deep learning enabled software MetronMind to measure vertebral heart size and vertebral left atrial size in dogs

Journal: bioRxiv
Published Date:

Abstract

Vertebral heart size (VHS) and vertebral left atrial size (VLAS) are objective radiographic measurements of heart and left atrial size respectively and are associated with inter and intraobserver variability when measured by humans. Artificial intelligence (AI) tools to determine VHS and VLAS have been developed which may reduce variability and save time. Two manual methods for measuring VHS and VLAS on right and left lateral canine thoracic radiographs were compared. Measurements of VHS and VLAS made by deep learning enabled program, MetronMind, were compared to a trained observer on right and left lateral radiographs from 1058 client-owned dogs including 80 breeds with a variety of heart sizes, thoracic conformations and radiographic quality. This was a retrospective, single center, method comparison study. Pearson’s correlation, Bland-Altman plots and Passing-Bablok regression were used to assess agreement. Correlation between traditional and modified manual measurements for VHS and VLAS were strong (r=0.994 and r=0.974 respectively), with minimal bias (−0.10 and 0.04 vertebrae respectively) indicating that the modified methods closely approximate traditional measurements obtained from right lateral views. MetronMind measurements of VHS and VLAS from right lateral radiographs correlated well with the human observer’s modified measurements (r=0.947 and r=0.811 respectively), showing small mean biases (0.08 and 0.07 vertebrae respectively). Correlation between left and right lateral radiographic measurements of VHS (0.87 and 0.91) was higher than for VLAS (0.73 and 0.64) and bias was larger for VHS (0.26 and 0.31 vertebrae) than VLAS (-0.13 and -0.10 vertebrae) for humans and MetronMind respectively. MetronMind can therefore assist veterinarians with measuring VHS and VLAS in dogs and right lateral radiographs are preferred. Future studies are needed to compare artificial intelligence derived radiographic measures with echocardiographic measures of cardiac size.

Authors

  • K. Tess Sykes; Sonya G. Gordon; John J. Craig; Sonya Wesselowski; Alice Watson