深度卷积神经网络集成在检测视网膜眼底照片中多种疾病方面比委员会认证眼科医生更准确可靠
研究开发了由五个卷积神经网络组成的深度卷积集成(DCE),用12个公开数据集的43055张眼底图像训练,将眼底照片分类为糖尿病视网膜病变(DR)、青光眼、年龄相关性黄斑变性(AMD)和正常眼。在100张未见测试图像上,DCE平均准确率79.2%,高于7名委员会认证眼科医生的72.7%(p=0.03),DR分类F1分数显著更高(76.8% vs 57.5%,p=0.01)。
要点
- DCE对四类眼底图像的平均准确率为79.2%,委员会认证眼科医生为72.7%(p=0.03)。
- DR分类F1分数DCE为76.8%,眼科医生为57.5%(p=0.01)。
- DCE准确率与置信度平均一致率为81.6%,眼科医生为70.3%(p<0.001)。
核心结论
- 深度卷积集成在四类眼底图像分类上比委员会认证眼科医生更准确可靠。
- 该工作提供原理验证,表明算法仅用眼底照片即可准确可靠识别多种视网膜疾病。
研究设计与样本
- 训练深度卷积集成(DCE),即五个卷积神经网络的集成,架构基于InceptionV3模型,初始权重在ImageNet数据集上预训练。
- 使用来自12个公开数据集的43055张眼底图像进行训练,将图像分类为糖尿病视网膜病变、青光眼、年龄相关性黄斑变性和正常眼。
- 五个训练好的集成在100张未见测试图像上测试,7名委员会认证眼科医生被要求分类这些测试图像。
主要发现
- 委员会认证眼科医生在所有类别上的平均准确率为72.7%,DCE为79.2%(p=0.03)。
- DR分类平均F1分数DCE为76.8%,眼科医生为57.5%(p=0.01)。
- 青光眼平均F1分数DCE为83.9%,眼科医生为75.7%(p=0.10),未达统计学显著。
- AMD平均F1分数DCE为85.9%,眼科医生为85.2%(p=0.69),未达统计学显著。
- 正常眼平均F1分数DCE为73.0%,眼科医生为70.5%(p=0.39),未达统计学显著。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
AIMS: To develop an algorithm to classify multiple retinal pathologies accurately and reliably from fundus photographs and to validate its performance against human experts. METHODS: We trained a deep convolutional ensemble (DCE), an ensemble of five convolutional neural networks (CNNs), to classify retinal fundus photographs into diabetic retinopathy (DR), glaucoma, age-related macular degeneration (AMD) and normal eyes. The CNN architecture was based on the InceptionV3 model, and initial weights were pretrained on the ImageNet dataset. We used 43 055 fundus images from 12 public datasets. Five trained ensembles were then tested on an 'unseen' set of 100 images. Seven board-certified ophthalmologists were asked to classify these test images. RESULTS: Board-certified ophthalmologists achieved a mean accuracy of 72.7% over all classes, while the DCE achieved a mean accuracy of 79.2% (p=0.03). The DCE had a statistically significant higher mean F1-score for DR classification compared with the ophthalmologists (76.8% vs 57.5%; p=0.01) and greater but statistically non-significant mean F1-scores for glaucoma (83.9% vs 75.7%; p=0.10), AMD (85.9% vs 85.2%; p=0.69) and normal eyes (73.0% vs 70.5%; p=0.39). The DCE had a greater mean agreement between accuracy and confident of 81.6% vs 70.3% (p<0.001). DISCUSSION: We developed a deep learning model and found that it could more accurately and reliably classify four categories of fundus images compared with board-certified ophthalmologists. This work provides proof-of-principle that an algorithm is capable of accurate and reliable recognition of multiple retinal diseases using only fundus photographs.
引用
- Pandey PU, Ballios BG, Christakis PG, et al. Ensemble of deep convolutional neural networks is more accurate and reliable than board-certified ophthalmologists at detecting multiple diseases in retinal fundus photographs. Br J Ophthalmol. 2024. PMID: 36720585. doi:10.1136/bjo-2022-322183
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。