What happened
A team at Hacettepe University, led by faculty member Olcay Hekimoğlu and supported by TÜBITAK, is developing an AI-powered mobile application designed to identify tick species from smartphone photographs. The project, which is scheduled to span three years, aims to provide rapid, preliminary identification of ticks, specifically focusing on species like Hyalomma marginatum that are known to transmit Crimean-Congo hemorrhagic fever (CCHF).
The project, titled 'A Smartphone-Based Artificial Intelligence Model for Rapid, Low-Cost and Accessible Identification of Tick Species in Türkiye,' is currently in development at Hacettepe University. The team is building a training database using reliable tick specimens that have been verified through both morphological characteristics and DNA analysis.
The AI model is being trained to recognize eight specific tick species, with a primary focus on Hyalomma marginatum, which is a known vector for CCHF. The application is designed to analyze photographs of ticks that have already been removed from a human or animal host, as the researchers noted that attached ticks may be obscured or physically altered, making identification difficult.
To ensure the tool is practical for real-world use, the team is testing the model across various smartphone devices and lighting conditions. The application is also being engineered to operate offline, which is a critical for users in rural or remote areas where internet access is not guaranteed.
The system includes a safety mechanism: if a submitted photograph is too blurry or insufficient for an accurate identification, the AI is programmed to request a new image or advise the user to seek expert evaluation rather than providing a forced, potentially incorrect result.
Source details: dailysabah.com ↗
Why it matters
This project addresses a significant public health challenge in Türkiye, where CCHF has resulted in over 17,000 cases and hundreds of deaths since 2002. By providing a low-cost, accessible tool for farmers, healthcare workers, and the general public, the researchers aim to improve early identification of disease-carrying ticks. The system is designed to function offline, ensuring utility in rural areas where internet connectivity is often limited. Furthermore, the project emphasizes real-world reliability by training the model on diverse, high-quality images and testing it under varied lighting and environmental conditions, rather than relying solely on laboratory-grade data. This approach could significantly enhance public awareness and data collection regarding tick distribution.
The primary public health goal is to provide a rapid, accessible preliminary identification tool for ticks, which are a persistent health threat in Türkiye. By enabling users to identify potential disease vectors quickly, the project aims to contribute to broader scientific data on tick distribution and public health safety.
The project distinguishes itself by prioritizing real-world performance over laboratory-only accuracy. By training the model on diverse, field-captured images and ensuring offline functionality, the researchers are addressing the practical limitations often found in AI-based diagnostic tools.
The researchers have clarified that the tool is not a substitute for professional medical diagnosis. Instead, it serves as an informational aid to help individuals and healthcare workers make more informed decisions about potential exposure to tick-borne diseases.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
In AI, what are a model's "parameters"?
What to watch next
The project is currently in a three-year development phase. Future milestones include the creation of a comprehensive, verified database of tick species using morphological and DNA-confirmed specimens, followed by rigorous testing in both controlled and field environments. The researchers have explicitly stated that the tool will not provide medical diagnoses, and it remains to be seen how the final application will be deployed to the public or integrated into existing health surveillance systems.
The project is planned for a three-year duration. During this time, the team will continue to refine the AI model and expand the database to include additional tick species in future versions.
Public availability and access conditions for the application have not yet been determined, as the project is still in the development and testing phase.
The effectiveness of the AI in real-world field conditions remains a key metric to monitor as the researchers transition from database creation to broader testing.