A groundbreaking machine learning algorithm developed by researchers from Middle Technical University in Iraq and the University of South Australia has achieved an impressive 98% accuracy in diagnosing diseases through the analysis of tongue color.

This AI imaging system represents a modern adaptation of traditional diagnostic practices, specifically those used in traditional Chinese medicine, which have been utilized for over 2,000 years.

The algorithm analyzes various features of the tongue, including color, shape, and thickness, to identify a range of health conditions such as diabetes, stroke, and COVID-19.

To develop this system, researchers trained the AI using a dataset of over 5,260 tongue images, which allowed for accurate predictions based on tongue coloration.

In testing, the model was validated with 60 images from patients at two teaching hospitals in the Middle East, successfully matching tongue color with the correct diagnoses nearly every time.

The diagnostic process involves capturing real-time images of the tongue using a webcam positioned about 20 centimeters away, which has proven effective in providing accurate diagnoses.

Despite its potential, challenges remain, including patient reluctance to share tongue images and issues with camera reflections that could affect diagnostic accuracy.

Looking ahead, researchers envision the integration of this technology into smartphones, enabling convenient, real-time health assessments for users, particularly in remote areas.

This innovation could significantly enhance early disease detection and management, democratizing healthcare and improving health outcomes on a global scale.

The findings of this research were published in the journal 'Technologies,' underscoring the potential of computerized tongue analysis in advancing medical diagnostics.

The research highlights the importance of integrating traditional medical practices with modern technology for more efficient disease screening.