Marzec Aleksandra, Filipowska Anna, Humeniuk Oliwia, Filipowski Wojciech, Raif Paweł
Foundation of Cardiac Surgery Development, Institute of Heart Prostheses, 345a Wolności, 41-800 Zabrze, Poland.
Department of Medical Informatics and Artificial Intelligence, Faculty of Biomedical Engineering, Silesian University of Technology, Roosevelta 40, 41-800 Zabrze, Poland.
Sensors (Basel). 2025 Aug 14;25(16):5038. doi: 10.3390/s25165038.
This article presents a deep learning approach for classifying the developmental stages (larvae, nymphs, adult females, and adult males) of ticks, the most common tick species in Europe and a major vector of tick-borne pathogens, including , , and tick-borne encephalitis virus (TBEV). Each developmental stage plays a different role in disease transmission, with nymphs considered the most epidemiologically relevant stage due to their small size and high prevalence. We developed a convolutional neural network (CNN) model trained on a dataset of microscopic tick images collected in the area of Upper Silesia, Poland. Grad-CAM, an XAI technique, was used to identify the regions of the image that most influenced the model's decisions. This work is the first to utilize a CNN model for the identification of European tick fauna stages. Compared to existing solutions focused on North American tick species, our model addresses the specific challenge of distinguishing developmental stages within . This solution has the potential to be a valuable tool in entomology, healthcare, and tick-borne disease management.
本文提出了一种深度学习方法，用于对蜱虫的发育阶段（幼虫、若虫、成年雌性和成年雄性）进行分类。蜱虫是欧洲最常见的蜱种，也是包括 、 和蜱传脑炎病毒（TBEV）在内的蜱传病原体的主要传播媒介。每个发育阶段在疾病传播中都起着不同的作用，由于若虫体型小且患病率高，被认为是流行病学上最相关的阶段。我们开发了一个卷积神经网络（CNN）模型，该模型基于在波兰上西里西亚地区收集的蜱虫微观图像数据集进行训练。使用一种可解释人工智能（XAI）技术Grad-CAM来识别对模型决策影响最大的图像区域。这项工作首次利用CNN模型来识别欧洲蜱类动物的发育阶段。与专注于北美蜱种的现有解决方案相比，我们的模型解决了区分 内发育阶段的特定挑战。该解决方案有可能成为昆虫学、医疗保健和蜱传疾病管理中的一个有价值的工具。
