arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.00197eess.IVcs.CVcs.LG
收藏

用于糖尿病视网膜病变严重程度预测的混合多模态深度ConvNet特征

Blended Multi-Modal Deep ConvNet Features for Diabetic Retinopathy Severity Prediction

Vignan's Foundation for Science Technology and Research(维格南科学技术研究基金会);University of Northern British Columbia(北不列颠哥伦比亚大学);Hankuk University of Foreign Studies(韩国外国语大学);VIT(韦洛尔理工学院);Chungbuk National University(忠北国立大学)

J. D. Bodapati, N. Veeranjaneyulu, S. N. Shareef, S. Hakak, M. Bilal, P. K. R. Maddikunta, O. Jo

摘要:

糖尿病视网膜病变(DR)是世界范围内视力损害和失明的主要原因之一。它通常见于长期患有糖尿病的患者。这项工作的主要重点是获得视网膜图像的最佳表示,进一步帮助提高DR识别模型的性能。为了提取最佳表示,从多个预训练的ConvNet模型中提取的特征使用所提出的多模态融合模块进行混合。这些最终表示用于训练用于DR识别和严重性级别预测的深度神经网络(DNN)。由于每个ConvNet提取不同的特征,因此使用1D池化和交叉池化融合它们比使用从单个ConvNet提取的特征更好地表示。对基准Kaggle APTOS 2019竞赛数据集的实验研究表明,在建议的混合特征表示上训练的模型优于现有方法。此外,我们注意到Xception和VGG16的基于交叉平均池的特征融合最适合DR识别。使用该模型,我们实现了97.41%的准确性和94.82的kappa统计量的DR识别和81.7%的准确性和71.1%的kappa统计量的严重程度预测。另一个有趣的观察结果是,与使用单峰深度特征训练的相同模型相比,在输入层具有dropout的DNN在使用混合特征训练时收敛得更快。

英文摘要:

Diabetic Retinopathy (DR) is one of the major causes of visual impairment and blindness across the world. It is usually found in patients who suffer from diabetes for a long period. The major focus of this work is to derive optimal representation of retinal images that further helps to improve the performance of DR recognition models. To extract optimal representation, features extracted from multiple pre-trained ConvNet models are blended using proposed multi-modal fusion module. These final representations are used to train a Deep Neural Network (DNN) used for DR identification and severity level prediction. As each ConvNet extracts different features, fusing them using 1D pooling and cross pooling leads to better representation than using features extracted from a single ConvNet. Experimental studies on benchmark Kaggle APTOS 2019 contest dataset reveals that the model trained on proposed blended feature representations is superior to the existing methods. In addition, we notice that cross average pooling based fusion of features from Xception and VGG16 is the most appropriate for DR recognition. With the proposed model, we achieve an accuracy of 97.41%, and a kappa statistic of 94.82 for DR identification and an accuracy of 81.7% and a kappa statistic of 71.1% for severity level prediction. Another interesting observation is that DNN with dropout at input layer converges more quickly when trained using blended features, compared to the same model trained using uni-modal deep features.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑