Computerized tongue image analysis for non-invasive disease screening: a review

Area of Science:

• Medical Imaging
• Biomarkers
• Digital Health

Background:

• Tongue surface and vein patterns are key in traditional diagnosis.
• Tongue imaging offers non-invasive insights into physiological and pathological variations.
• Conventional tongue diagnosis relies on subjective judgment, impacting accuracy.

Purpose of the Study:

• To propose a structured taxonomy for computerized tongue image analysis (CTIA).
• To enhance objectivity, reproducibility, and consistency in tongue-based diagnostics.
• To provide a framework for integrating tongue imaging into clinical practice for disease screening.

Main Methods:

• Systematic review of advances in CTIA stages: image acquisition, preprocessing, dataset construction, feature extraction, and disease detection.
• Development of a structured taxonomy for CTIA.
• Identification of challenges and solutions in data standardization and feature quantification.

Main Results:

• A comprehensive taxonomy for CTIA is proposed.
• Key challenges in CTIA, such as data standardization and feature quantification, are identified.
• Potential solutions to improve diagnostic precision and reliability are outlined.

Conclusions:

• CTIA offers a path to more objective and reliable health assessment.
• The proposed taxonomy can guide the development and clinical integration of tongue imaging tools.
• Tongue imaging, supported by CTIA, shows promise for non-invasive disease screening and continuous health monitoring.