Oren, A., Türkcü, J. D., Meller, S., Lazebnik, T., Wiegel, P., Mach, R., Volk, H. A., & Zamansky, A. (2023). BrachySound: machine learning based assessment of respiratory sounds in dogs. Scientific Reports, 13, Article 20300.
This study investigates the application of machine learning models to improve the diagnosis of brachycephalic obstructive airway syndrome (BOAS) in dogs. By analyzing 366 audio samples from 69 Pugs and 79 other brachycephalic breeds recorded during a 15-minute standardized exercise test, the models achieved a peak accuracy of 0.85 in classifying BOAS test results. Predictions from rest recordings yielded accuracies of 0.68 for Pugs and 0.65 for various brachycephalic breeds, with an F1 score of 0.80 for detecting laryngeal sounds. The findings suggest that machine learning could enhance objectivity and efficiency in BOAS diagnostics compared to traditional methods.
Comments