Development of an AI-Based Smartphone Application for Rapid Tick Identification and Geospatial Mapping: A Pilot Study

Tick-borne diseases, particularly spotted fever group rickettsioses in the western regions of Japan, represent an increasing public health concern. Accurate tick identification is clinically important because pathogen transmission varies by tick genus. However, tick identification is rarely performed in routine clinical practice. Between May 2023 and December 2024, ticks were collected from patients presenting tick bites at 10 medical institutions in Okayama, Hiroshima, and Kagawa prefectures. Additional images of wild ticks were included in the dataset to address class imbalances. Trained personnel morphologically identified all ticks. We developed a two-stage artificial intelligence (AI) pipeline comprising object detection and image classification. Tick regions were detected using a YOLO-based model trained on a publicly available dataset (3,258 annotated images), followed by genus-level classification using deep learning models. The system was integrated into a prototype smartphone application with geospatial visualization. The object detection model achieved a mean average precision (mAP@0.5) of 0.954, demonstrating highly accurate tick localization. For classification, a curated dataset of 533 tick images representing four genera was used. The ResNet50 model achieved a mean validation accuracy of 95%, despite substantial class imbalance. Visualization analyses confirmed that the model primarily focused on the morphological features of ticks rather than background elements. We developed the first AI-based smartphone application in Japan for automated tick identification and geospatial visualization. Although limited to genus-level identification and microscope-acquired imaging, this approach may support clinical risk assessment, facilitate patient counseling, reduce unnecessary empirical antibiotic use, and enhance public health surveillance.