Re: “Correlation of Automated Computed Tomography Volumetric Analysis Metrics With Motility Disturbances in Thyroid Eye Disease”

Abstract

Letters to the editorRe: "Correlation of Automated Computed Tomography Volumetric Analysis Metrics With Motility Disturbances in Thyroid Eye Disease" To the Editor:We read with great interest the article "Correlation of Automated Computed Tomography Volumetric Analysis Metrics With Motility Disturbances in Thyroid Eye Disease" wherein Law et al. 1 employ an automated segmentation method developed using normal orbital CT images to characterize relationships between CT metrics and clinical findings in patients with thyroid eye disease. Using this automated method, the authors concluded that ocular motility deficits were correlated with superior, inferior, and lateral recti average diameter, maximum diameter, and volume, and medial rectus volume. 1 Developing a fully automated method of segmenting orbital soft tissues to help analyze patients with thyroid eye disease holds great potential value. However, any new method must be validated for accuracy. As we have presented at the 2022 ASOPRS Fall Meeting, our fully-automated atlas-based method developed using normal control orbits demonstrated lower accuracy in evaluating soft tissue volumes in thyroid eye disease orbits compared with normal orbits, using manual segmentation as the gold standard. 2 Accuracy was assessed using intersection over union and Dice coefficient, which is standard for analyzing such imaging algorithms. 2,3 As the automated segmentation method described by Law et al. 1 was developed using normal orbit CT images, further validation must be presented to describe the accuracy of this new method as applied to thyroid eye disease orbits before clinical associations can be meaningfully analyzed.We thank the authors for bringing this important new area of research to the oculofacial literature.