Once upon a time, in a kingdom not too far away, there was a doctor who always asked his patients to show their tongues. Today, that same request could come from your smartphone. Welcome to the future of medical screening, where an AI inspired by age-old practices can diagnose your health conditions with an accuracy that would make even the most expert of doctors pale. Now, before you “click” on your tongue, read here.
A fusion of ancient and modern
An international team of researchers has developed an artificial intelligence algorithm that promises to revolutionize the way we conduct medical screening (you can find the study here). This innovation is based on a 2000-year-old practice of traditional Chinese medicine: tongue analysis.
Ali Al-Naji, adjunct professor at Baghdad Central Technical University and the University of South Australia, explains:
The color, shape and thickness of the tongue can reveal a multitude of health conditions.
The proposed system automates this ancient medical practice of lingual diagnosis with the help of modern machine learning. The researchers powered the algorithm with over 5.200 language photos (a little bit, personal opinion) and the corresponding diagnoses, using six different machine learning algorithms.
The result? Surprising accuracy in detecting health conditions based solely on tongue color.
Head-spinning precision. Or rather, the language
In a test with 60 photos of languages from teaching hospitals in the Middle East, the AI was able to accurately detect each patient's condition 98% of the time. Conditions ranged from gastrointestinal issues to COVID-19.
Even the least accurate system, known as methodology naïve Bayes, showed an accuracy rate above 91%. The open-source framework XGBoost He has reached an incredible accuracy of 98,71%.
The ultimate goal of this research team is to bring this technology directly into the hands of consumers. Perhaps creating a smartphone app that uses this algorithm, allowing people to get predictive diagnoses from the comfort of their own home.
Implications for the future of medical screening
This innovation opens up interesting scenarios on the future of medical screening:
- Accessibility: Could it make basic screening more accessible in areas with limited medical resources?
- Prevention: Could it encourage more frequent checks, leading to earlier diagnoses?
- Doctor workload: How might it influence the role of doctors in primary diagnosis?
And clearly, there are just as many unknowns to consider:
- Data protection: How will sensitive health information be managed and protected?
- Overdiagnosis: There is a risk of causing unnecessary anxiety (a sort ofdoomscrolling” healthcare) through false positives?
- Technological addiction: Could it lead to a decrease in traditional diagnostic skills?
Even with all the potential it has, we must be cautious in implementing this technology. Medical diagnosis is a complex process that often requires more than a single indicator.
Is the future of screening a chorus of digital “ahhhs”?
The idea of screening with tongue selfies may seem bizarre, but this technology has the potential to democratize access to primary care medical screening, especially in areas with limited resources.
The future of medical screening may be closer than we think, and it could start with a simple "cheese" to our smartphone. But remember: even if AI can read your language, it is always better to consult a real doctor for a complete diagnosis. After all, no algorithm can replace the experience and intuition of a medical professional… at least for now.
