Advanced Computer Vision Techniques for Automated Identification of Lyme Disease-Vector Ticks

The infectious illness Lyme disease results from Borrelia burgdorferi which spreads between humans when blacklegged ticks (Ixodes scapularis) transmit microorganisms to eastern regions of North America. Medical experts recommend taking antibiotics as prophylaxis for Lyme disease prevention when such medication is given within 72 hours after a tick bite. The identification process for blacklegged ticks needs to be precise for people to receive prompt medical intervention. Researchers developed an automatic system that uses current computer vision technology for real-time tick types classification. This study trains an advanced deep learning Convolutional Neural Networks ResNet system as part of a complete end-to-end process to identify blacklegged ticks among other tick species. The model performance should be enhanced through implementing sophisticated transfer learning methods. The most successful CNN ResNet model delivers an outstanding accuracy rate of $\mathbf{9 8. 2 6 \%}$ for detecting unidentified tick species during testing. The automated detection method simplifies tick identification in addition to supporting public health monitoring efforts regarding tick-borne disease surveillance. Such technology shows promise to unite with geographical information to enhance Lyme disease risk analysis capabilities. This research stands as the initial deep learning application for tick sorting which leads to enhanced automation of tick observation operations while opening new possibilities to advance tick disease control frameworks as well as ecological study methodologies.