摘要
目的:慢性伤口的图像分割是伤口护理的关键之一,不同的损伤程度与周围皮肤会增加分割的难度,本研究构建基于人工智能算法的足部慢性伤口图像快速、准确的分割与识别方法,用于辅助伤口护理。方法:首先对图像进行标注,然后使用深度学习进行图像分割训练。训练了两种数据集,其中足部溃疡分割挑战使用交并比(IOU)相关指标进行评估,Medetec足溃疡-224数据集使用DICE系数进行评估。结果:使用该方法进行的图像分割训练,获得的图片结果较为准确。对于两种数据集,无论是训练集还是验证集均具有优异的准确度。结论:本文构建的分割方法可用于足部慢性伤口的分割实现图像的精确识别,取得了令人满意的分割效果,在辅助足部慢性伤口护理中的应用与提高护理质量方面具有一定的实用价值与借鉴意义。
Objective:Image segmentation of chronic wounds is one of the keys to wound care,and different degrees of injury and surrounding skin will increase the difficulty of segmentation.In this study,a fast and accurate segmentation and recognition method for chronic foot wounds based on artificial intelligence algorithm was constructed to assist wound care.Methods:The images were labeled firstly,and then deep learning was used for image segmentation training.Two datasets were trained,in which the foot ulcer segmentation challenge was evaluated using intersection-ratio(IOU)related indicators,and the Medetec foot ulcer-224 dataset was evaluated using DICE coefficient.Results:The image segmentation training with this method obtained more accurate.For both data sets,both the training set and the validation set have excellent accuracy.Conclusion:The segmentation method constructed in this paper can be used for the segmentation of chronic foot wounds to achieve accurate image recognition,and has achieved satisfactory segmentation effect,which has certain practical value and reference significance in assisting the application of chronic foot wound nursing and improving the quality of nursing.
出处
《黑龙江中医药》
2023年第5期148-150,共3页
Heilongjiang Journal of Traditional Chinese Medicine
