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: ($D{S}_1$) a simpler system based on very fast handcrafted image feature extraction and support vector machine classification and ($D{S}_2$) a more complex system based on a deep neural network, Vgg-16, for image feature extraction and classification. $D{S}_1$ is much faster than $D{S}_2$, but it is less accurate than $D{S}_2$. Fortunately, the errors of $D{S}_1$ are not the same of $D{S}_2$. During training, we use a validation set to learn the probabilities of misclassification by $D{S}_1$ on each class based on its confidence values. When $D{S}_1$ 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 $D{S}_2$. 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.

肠道寄生虫是人类几种疾病的原因。为了消除光学显微镜载玻片易于出错的视觉分析，我们研究了用于诊断人类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属性的决策系统的意见：$d{\mathrm{小号}}_{\mathrm{1个}}$）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（$d{\mathrm{小号}}_2$）基于深度神经网络Vgg-16的更复杂的系统，用于图像特征提取和分类。 $d{\mathrm{小号}}_{\mathrm{1个}}$ 比 $d{\mathrm{小号}}_2$，但准确性不如 $d{\mathrm{小号}}_2$。幸运的是，$d{\mathrm{小号}}_{\mathrm{1个}}$ 与...不同 $d{\mathrm{小号}}_2$。在训练期间，我们使用验证集来了解错误分类的可能性$d{\mathrm{小号}}_{\mathrm{1个}}$在每个类别上根据其置信度值。什么时候$d{\mathrm{小号}}_{\mathrm{1个}}$ 快速地从显微镜载玻片上对所有图像进行分类，该方法选择了许多错误分类几率较高的图像，以进行表征和重新分类 $d{\mathrm{小号}}_2$。我们的混合系统可以在不影响效率的情况下提高整体效率，适合临床常规操作-一种可能适合其他实际应用的策略。如大型数据集所示，该系统在蠕虫卵，蠕虫幼虫和原生动物囊肿上平均可分别达到Cohen's Kappa的94.9％，87.8％和92.5％。