摘要
为了构建可解释性AI模型以实现对COVID-19进行快速准确诊断,本文基于深度学习算法的图像分类技术,通过对胸片数据集的预处理和数据增强,运用Python编程语言和Pytorch深度学习框架构建了CNN和DenseNet模型,实现对肺部疾病的自动诊断,并对两个模型的性能进行了评价和对比分析。结果表明,DenseNet模型在图像分类任务上表现更优,具有更好的分类准确率和更快的收敛速度。本研究为医疗图像识别领域的研究提供了新的思路和方法,对于提升医学诊断的准确性和效率具有一定的应用价值。
In order to build an interpretable AI model to realize rapid and accurate diagnosis of COVID-19, this paper uses the Python programming language and Pytorch deep learning framework to build CNN and DenseNet models based on the image classification technology of deep learning algorithm and through the preprocessing and data enhancement of chest X-ray data sets. The automatic diagnosis of lung diseases is realized, and the performance of the two models is evaluated and compared. The results show that the DenseNet model performs better in image classification task, has better clas-sification accuracy and faster convergence rate. This study provides a new idea and method for the research in the field of medical image recognition and has certain application value for improving the accuracy and efficiency of medical diagnosis.
出处
《建模与仿真》
2023年第3期3024-3034,共11页
Modeling and Simulation
