Back

Accuracy of computed tomography in diagnosing temporomandibular joint osteoarthritis relative to histopathological findings - an ex-vivo study of 41 horses

Jasinski, T.; Zdrojkowski, L.; Turek, B.; Kaczorowski, M.; Pawlinski, B.; Brehm, W.; Domino, M.

2025-10-27 pathology
10.1101/2024.11.01.621628 bioRxiv
Show abstract

Computed tomography (CT) is commonly used for imaging the equine head, particularly when diseases of the temporomandibular joint (TMJ), including osteoarthritis (OA), are suspected. However, the accuracy of CT-based diagnosis of TMJ OA remains uncertain. This prospective descriptive cross-sectional study aimed to describe the CT and histopathological findings of TMJ OA concerning horse age, assess the correlation between the CT and histopathological findings, and evaluate the accuracy of CT-based diagnosis of TMJ OA. CT images and tissue samples from 41 heads (82 TMJs) were independently and simultaneously reviewed by two observers. Descriptive statistics on both findings, their age- and TMJ OA- related frequency distribution, and the accuracy of TMJ OA diagnosis were calculated. Based on CT findings, TMJ OA was recognized in 50% of joints in horses aged 1-4 years and 100% in horses aged 5-15 and >15 years. However, based on histopathological findings, TMJ OA was confirmed in 18% of joints in horses aged 1-4 years, 63% in horses aged 5-15 years, and 83% in horses >15 years. The occurrence of both findings differed among horses of different ages. Consequently, the CT-based diagnosis of TMJ OA showed moderate accuracy with a 0.79 sensitivity and 0.50 specificity. However, accuracy metrics also differed among horses of different ages. Equine TMJ may undergo both age- and disease-related remodeling, detectable through CT and histopathology. However, the sensitivity and specificity of CT- based TMJ OA diagnosis may be affected by the consideration of CT anatomical variations.

Published in Animals (predicted rank #3) · training set

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.