摘要
糖尿病视网膜病变(DR)是一种常见的眼科疾病，可导致视力问题和失明，需要准确分级以有效治疗。虽然已经开发了各种人工智能(AI)系统，在检测DR方面超过了人类分析，但深层神经网络需要大型注释数据集来学习评分所需的复杂模式和关系，这些模式和关系往往在可用性上受到限制，以学习准确评分所需的复杂模式和关系。然而，这种数据集的可用性往往有限，需要大量的人力资源投资和标记过程的时间。为了应对这些挑战，我们提出了ESSP-CNS，这是一个利用流行的CNN架构（VGGNet、AlexNet和ResNet）的框架。我们的方法采用了自我监督学习，特别是引导你自己潜在的(BYOL)技术，在一个巨大的未标记数据集上预训练神经网络。此外，我们采用深度集成学习方法构建了一个稳健的DR评分模型。我们的方法包括三个主要部分：眼底图像预处理、基于byol的预训练和集成模型构建。我们使用EyePACS和IDRiD数据集进行实验和比较，在EyePACS上使用BYOL预训练，使CNN模型获得有意义的眼底图像表示，而IDRiD用于严重程度分级。通过Messidor数据集的验证进一步验证了该框架的性能。在IDRiD和MESIDOR数据集上的大量实验表明,ESP-CNN的准确率分别为71.84%和75.42%,特异性分别为88.76%和87.13%,AUC分别为86.02%和86.54%。实验结果验证了我们的方法在DR严重程度分级中的有效性，由预先训练的CNN建立的集成模型产生了良好的结果。此外，我们将我们的方法与其他先进的DR分级方法进行了比较，我们的结果表明它的性能令人满意，在准确评估DR严重程度方面超过了以前的替代方法。
Abstract
Diabetic retinopathy (DR) is a common eye disorder that can lead to vision problems and blindness, necessitating accurate grading for efective treatment. While various artifcial intelligence (AI) systems have been developed, surpassing human analysis in detecting DR, deep neural networks require large annotated datasets to learn the complex patterns and relationships necessary for grading, which are often limited in availability, to learn the intricate patterns and relationships required for accurate grading. However, such data- sets are often limited in availability, requiring signifcant investments of human resources and time for the labeling process. To address these challenges, we propose ESSP-CNNs, a framework that harnesses popular CNN architectures (VGGNet, AlexNet, and ResNet). Our approach employs self-supervised learning, specifcally the Bootstrap Your Own Latent (BYOL) technique, to pre-train neural networks on a vast unlabeled dataset. Additionally, we employ deep ensemble learning to construct a robust model for DR grading. Our meth- odology encompasses three main components: preprocessing fundus images, BYOL-based pre-training, and ensemble model construction. We conduct experiments and comparisons using the EyePACS and IDRiD datasets, with BYOL pre-training on EyePACS to enable the CNN models to acquire meaningful representations of fundus images, while IDRiD is used for severity grading. The performance of the proposed framework is further con- frmed through thorough validation using the Messidor dataset. Through extensive experi- mentation on the IDRiD and Messidor datasets, ESSP-CNNs achieve notable accuracies of 71.84% and 75.42%, specifcities of 88.76% and 87.13% along with AUC of 86.02% and 86.54%, respectively. The experimental results validate the efectiveness of our methodol- ogy in grading the severity of DR, with the ensemble model built from pre-trained CNNs yielding promising outcomes. Moreover, we compare our methodology against other state- of-the-art methods in DR grading, and our results demonstrate its satisfactory performance, surpassing previous alternatives in accurately assessing DR severity.
NETL
NSTL