Coming out of Iraq and Australia, scientists have developed an AI where the algorithm is capable of diagnosing diseases based on the colour of a person’s tongue. This new thirty developed by Xu et al. , and having an extraordinary 98% accuracy, might usher in a new era of medical diagnostics which is easier, quicker and more available to millions of people.
AI Algorithm Science
TCM & Modern Technology
The AI system is inspired by the practice of TCM, where tongue examination has been used for many years to determine different diseases. In the TCM the color, form and coating of the tongue are believed to be an indication of health of an individual. The researchers thought about this old philosophy and duly designed a modern AI model which has the ability to study these characteristics and diagnose diseases.
How the Technology Works
Data Collection: These three databases consisted of 5,260 images of tongues extracted from patients attending two teaching hospitals in the Middle East. These images gave a wide-ranging dataset that included various diseases and the-specific feature of the tongue in question.
Image Analysis: The patients were placed at approximately 8 inches from a special laptop with an average webcam. Using a webcam, the physical state of the patient’s tongue was photographed and then introduced to the AI algorithm’s examination.
Color and Condition Correlation: The algorithm locates certain shades and patterns on the surface of the tongue which are connected to specific illnesses. For instance:
Yellow Tongue: Sometimes people found out they had it when they were diagnosed with, or suspected to have, diabetes.
Purple Tongue with Thick Greasy Coating: Don’t exactly point to cancer, but make you suspicious enough to consult your doctor and get tested.
Unusually Shaped Red Tongue: Pathogenic manifestation of acute stroke.
White Tongue: May be an indication of anemia.
Deep Red Tongue: Correlate with high risk of developing clinical manifestations of COVID-19.
Indigo or Violet Tongue: Can raise concerns about vascular or gastrointestinal disease, or asthma.
Results and Accuracy
There are 60 tongue images used to test the performance of this AI model; the result for this model achieved 98% accuracy for diagnosis diseases corresponded to the tongue image. Such high accuracy of tongue recognition serves as a basis for suggesting the primary potential of AI application for tongue diagnostics.
You May Also Read: Yamaha’s RW Series: A New Line of Motorcycles on the Horizon
Potential Applications and Benefits
A Future in Your Pocket: Smartphone Integration:
Another potential of this technology is the idea of incorporating of this technology into a smartphone application. It has been observed that an app of this nature would enable its users to diagnose themselves to diseases such as diabetes, stroke, anaemia, asthma, liver and gallbladder complaints, and even COVID-19 without necessarily having to present themselves physically to a healthcare professional. This would go a long way in minimizing the frequent hospital attendance especially in areas with few if any health facilities.
I have identified the following benefits arising from the AI tongue analysis:
Non-Invasive and Quick: Competitive to previous diagnostic techniques which could call for a blood test or any invasive procedure, the current AI system only needs the picture of the tongue.
User-Friendly and Accessible: The aim of the system design is to make it user-friendly so that it does not require a highly technical person to use it. This makes its use possible by a vast number of users especially the people in the rural areas who sometimes cannot easily access the health facilities.
Cost-Effective: This technology could have the potential of saving much money to the patients and the hospital since multiple tests and hospital visits would be eliminated.
Supports Modern Medicine with Ancient Practices: The application of this AI is highly effective in confirming the current diagnostic techniques and confirming the usefulness of the traditions identified in TCM.
Challenges and Considerations
Addressing Data Privacy Concerns
All the same, the application of the AI tongue analysis comes with a number of challenges. A limitation, for example, is the amount of resistance patients show to disclose their personal information or pictures of their tongue. To overcome this challenge, the developers will have to make sure that the system is secure or otherwise patient information will be compromised.
Technical Hurdles: To avoid the above mentioned camera reflections, one needs to remember the following tips:
Another difficulty is that some of the reflections can be observed by the camera and, therefore, can deceive the algorithm. Mirrors tarnish the image of tongue and cause misleading conclusions of the disease. In an effort to resolve this, the researchers are most probably going to develop increasingly effective methods to eliminate such distortions.
The paper calls for further validation and testing of the model.
Although they show a good perspective for the using, additional trials and verifications are needed to promote the technology. Higher volumes of different types of data will be necessary to ensure that the AI system will be able to correctly diagnose not only many more diseases, but across different demos.
The four emerging uses of AI in healthcare are as follows:
The advancement of this AI tongue analysis technology is just one example of the implementing of AI in the healthcare sector. It is possible to assume that more brilliant solutions that contribute to the increase in the accuracy, speed, and availability of diagnosis will appear as AI develops.
The Path Forward
Scholars are bullish on this technology. Another opinion belongs to the study co-author Javaan Chahl who reasoned that computerized tongue analysis will eventually be mainstream as a disease screening determinant. He opines that it is secure, efficient, easy to use, and cheap and integrates technology with one of the oldest forms of practice.
Implications for Global Health
It can be applied at regional and global levels hence has high potential of being implemented. An application that can diagnose severe health conditions for instance in areas where individuals cannot afford to seek medical attention can be a lifesaver. Also, with the increasing use of AI-based diagnostics, such solutions can partly offset the pressure in healthcare, since more and more services will be able to dispense with personal consultations and examinations.
Conclusion
The recent creation of an AI model that can diagnose diseases with a 98% efficiency, by merely considering the colour of one’s tongue, is a breakthrough in medical science. Thus, by integrating the principles of the ancient tongue examination practice into the AI approach, the researchers have provided people with the opportunity to use a service that is accurate and effective at a rather low price. Despite the shortcomings that can still be observed, the opportunities that can be achieved through this technology are enormous and the technology may in the near future become an important aspect of health care delivery system.
Reference(AI model 98% accurate in detecting diseases — just by looking at your tongue)