Osaku D, Cuba C F, Suzuki C T N, Gomes J F, Falcão A X
Institute of Computing, University of Campinas, Brazil.
Comput Biol Med. 2020 Aug;123:103917. doi: 10.1016/j.compbiomed.2020.103917. Epub 2020 Jul 15.
Intestinal parasites are responsible for several diseases in human beings. In order to eliminate the error-prone visual analysis of optical microscopy slides, we have investigated automated, fast, and low-cost systems for the diagnosis of human intestinal parasites. In this work, we present a hybrid approach that combines the opinion of two decision-making systems with complementary properties: (DS) a simpler system based on very fast handcrafted image feature extraction and support vector machine classification and (DS) a more complex system based on a deep neural network, Vgg-16, for image feature extraction and classification. DS is much faster than DS, but it is less accurate than DS. Fortunately, the errors of DS are not the same of DS. During training, we use a validation set to learn the probabilities of misclassification by DS on each class based on its confidence values. When DS quickly classifies all images from a microscopy slide, the method selects a number of images with higher chances of misclassification for characterization and reclassification by DS. Our hybrid system can improve the overall effectiveness without compromising efficiency, being suitable for the clinical routine - a strategy that might be suitable for other real applications. As demonstrated on large datasets, the proposed system can achieve, on average, 94.9%, 87.8%, and 92.5% of Cohen's Kappa on helminth eggs, helminth larvae, and protozoa cysts, respectively.
肠道寄生虫会引发人类的多种疾病。为了消除光学显微镜载玻片易出错的视觉分析，我们研究了用于诊断人体肠道寄生虫的自动化、快速且低成本的系统。在这项工作中，我们提出了一种混合方法，该方法结合了两个具有互补特性的决策系统的意见：（DS）一个基于非常快速的手工图像特征提取和支持向量机分类的更简单系统，以及（DS）一个基于深度神经网络Vgg - 16进行图像特征提取和分类的更复杂系统。DS比DS快得多，但准确性不如DS。幸运的是，DS的错误与DS的不同。在训练期间，我们使用一个验证集，根据DS的置信度值来了解其在每个类别上误分类的概率。当DS快速对显微镜载玻片上的所有图像进行分类时，该方法会选择一些误分类可能性较高的图像，由DS进行特征描述和重新分类。我们的混合系统可以在不影响效率的情况下提高整体有效性，适用于临床常规——这种策略可能适用于其他实际应用。如在大型数据集上所展示的，所提出的系统在蛔虫卵、蛔虫幼虫和原生动物囊肿方面，平均分别可达到科恩卡帕系数的94.9%、87.8%和92.5%。
