Coronary artery disease (CAD) remains the leading cause of death globally, responsible for approximately 17.9 million deaths each year, according to the World Health Organization.
Currently, the most reliable method for diagnosing CAD is coronary angiography, a procedure that, while effective, is costly, invasive, and not suitable for early diagnosis or risk assessment.
In an effort to find a safer, more affordable, and efficient diagnostic method, researchers from Beijing University of Chinese Medicine and Hunan University of Chinese Medicine have turned to artificial intelligence (AI) and traditional Chinese medicine (TCM).
Their innovative approach, which utilizes AI to analyze tongue images, has been detailed in a recent publication in Frontiers in Cardiovascular Medicine.
Traditional Chinese medicine has long held that “internal diseases manifest externally,” meaning that outward signs and symptoms, such as those observed on the tongue, can provide valuable insights into a person’s health.
The tongue, rich in nerves and blood vessels, is closely linked to the cardiovascular system, making it a useful indicator of systemic health conditions, particularly those affecting blood circulation.
This ancient practice of tongue diagnosis, which involves examining the color, coating, and shape of the tongue, has been shown in numerous studies to be an effective tool for diagnosing various diseases.
Building on this tradition, the research team set out to determine whether tongue images could be used as a non-invasive method to diagnose CAD. They developed a machine learning model that combines traditional risk factors for CAD with deep features extracted from tongue images.
The model was built using ResNet-18, a deep learning network pre-trained on ImageNet, and further refined using XGBoost, an algorithm that provided the best classification results when integrating both risk factors and tongue image features.
To test their model, the researchers recruited 410 patients from four hospitals between March 2019 and November 2022. These participants, aged between 18 and 85, were either diagnosed with hypertension or a combination of hypertension and CAD.
The model’s performance was particularly strong in individuals aged 65 and older and was equally effective for both men and women. This suggests that the AI-driven tongue imaging method has a broad applicability across different demographic groups.
One of the key findings of the study was that the inclusion of tongue image features significantly enhanced the accuracy of the CAD diagnosis compared to models that relied solely on traditional risk factors.
The diagnostic model demonstrated an overall accuracy of greater than 99% when utilizing tongue images alongside traditional risk factors.
This highlights the potential of incorporating tongue analysis into modern diagnostic techniques, blending the wisdom of TCM with cutting-edge AI technology.
However, the study was not without its limitations. The sample size was relatively small and only included patients with hypertension, which may limit the generalizability of the findings.
Additionally, the study only used a single type of equipment to capture tongue images, which could restrict the applicability of the diagnostic model in different settings or with different devices.
The researchers acknowledged these limitations and emphasized the need for future studies with larger and more diverse populations to validate and optimize the diagnostic model.
Looking ahead, the research team suggests that integrating additional biomarkers with tongue images could further enhance the diagnostic accuracy of their model.
Despite the limitations, their work represents a promising step forward in the development of non-invasive, AI-driven diagnostic tools for CAD.
By combining the insights of traditional Chinese medicine with modern technology, the researchers have introduced a novel approach that could one day offer a simpler, more accessible means of diagnosing one of the world’s deadliest diseases.
In conclusion, this study demonstrates the feasibility of using tongue images as a diagnostic tool for CAD, potentially offering a new risk indicator for the disease.
As AI technology continues to evolve, it may open up new avenues for integrating ancient medical practices with contemporary healthcare, providing patients with more effective and less invasive diagnostic options.
If you care about health, please read studies about the benefits of low-dose lithium supplements, and what we know about egg intake and heart disease.
For more health information, please see recent studies about potatoes and high blood pressure, and results showing 6 best breads for people with heart disease.
The research findings can be found in Frontiers in Cardiovascular Medicine.
Copyright © 2024 Knowridge Science Report. All rights reserved.
