Brachycephalic Dog Nostril Images for Stenosis Degree Classification

本研究首次提出了一种针对短头犬种鼻孔狭窄程度分类的新型图像数据集，包含190张法国斗牛犬的鼻孔图像。数据集由多姆博斯科天主教大学与OdontoPet诊所合作，从2021年7月至2023年4月收集，涵盖了95只不同性别、年龄和体型的法国斗牛犬。数据集详细记录了鼻孔的狭窄程度，分为轻度、中度和重度，以及非狭窄状态。该数据集旨在通过深度学习技术，自动识别和分类鼻孔狭窄程度，以辅助兽医和宠物主人在诊断和治疗短头犬呼吸障碍方面提供科学依据。

This study proposes for the first time a novel image dataset for classifying the severity of nostril stenosis in brachycephalic dog breeds, which includes 190 nostril images of French Bulldogs. The dataset was collected between July 2021 and April 2023 through a collaboration between Dom Bosco Catholic University and OdontoPet Clinic, and involves 95 French Bulldogs with diverse genders, ages and body types. The dataset comprehensively records the severity of nostril stenosis, which is categorized into four grades: mild, moderate, severe and non-stenotic. This dataset aims to facilitate automatic identification and classification of nostril stenosis severity via deep learning technologies, so as to provide scientific basis for assisting veterinarians and pet owners in the diagnosis and treatment of respiratory disorders in brachycephalic dogs.