Identification of public submitted tick images: A neural network approach

Abstract

Ticks and tick-borne diseases represent a growing public health threat in North America and Europe. The number of ticks, their geographical distribution, and the incidence of tick-borne diseases, like Lyme disease, are all on the rise. Accurate, real-time tick-image identification through a smartphone app or similar platform could help mitigate this threat by informing users of the risks associated with encountered ticks and by providing researchers and public health agencies with additional data on tick activity and geographic range. Here we outline the requirements for such a system, present a model that meets those requirements, and discuss remaining challenges and frontiers in automated tick identification. We compiled a user-generated dataset of more than 12,000 images of the three most common tick species found on humans in the U.S.: Amblyomma americanum, Dermacentor variabilis, and Ixodes scapularis. We used image augmentation to further increase the size of our dataset to more than 90,000 images. Here we report the development and validation of a convolutional neural network which we call “TickIDNet,” that scores an 87.8% identification accuracy across all three species, outperforming the accuracy of identifications done by a member of the general public or healthcare professionals. However, the model fails to match the performance of experts with formal entomological training. We find that image quality, particularly the size of the tick in the image (measured in pixels), plays a significant role in the network’s ability to correctly identify an image: images where the tick is small are less likely to be correctly identified because of the small object detection problem in deep learning. TickIDNet’s performance can be increased by using confidence thresholds to introduce an “unsure” class and building image submission pipelines that encourage better quality photos. Our findings suggest that deep learning represents a promising frontier for tick identification that should be further explored and deployed as part of the toolkit for addressing the public health consequences of tick-borne diseases.

Lyme rashes disease classification using deep feature fusion technique

“…For early Lyme disease identification based on gene expression, Servellita et al31 created a diagnostic classifier with a 95.2% accuracy. Justin et al66 trained a CNN to identify tick bites using a photo dataset collected via a mobile app.…”

Monitoring Trends in Distribution and Seasonality of Medically Important Ticks in North America Using Online Crowdsourced Records from iNaturalist

“…Whilst iNaturalist datasets covering similar time periods are comparably smaller, the rapid year-on-year increases in tick observations by the community suggest that this dataset will continue to grow. Although slightly quicker than the morphological identification of tick samples due to the time associated with specimen handling, the identification of ticks via photographs is still a time-consuming process, and advances in automated tick identification [ 75 ] could significantly improve speed in the future. To gain a rapid overview of tick data, it is also possible to rely on the iNaturalist community identifications, which are generally good, although less common species may be misidentified; for example, during this study it was noticed that A. maculatum observations were sometimes misidentified as Dermacentor , particularly in areas where the Gulf Coast tick is less common.…”

Evaluating spatial and temporal patterns of tick encounters using community science data submitted through a smartphone application

“…Although this method is ultimately sensitive to the image quality and the level of experience of the individual who is responsible for identifying the tick, identification can be verified by trained entomologists (Fernandez et al, 2019). Furthermore, machine learning methods are increasingly being used to expediate identification processes and minimize biases (Justen et al, 2021;Kopsco et al, 2021a).…”