Introducing a computer-assisted, digital tongue-imaging device for standardization of traditional tongue diagnosis.

Background: In Traditional Chinese Medicine (TCM) tongue diagnosis and pulse diagnosis are the major diagnostic methods up till now. As western research has tended to focus on acupuncture, attempts to standardize the classic diagnostic methods have been few.

Methods: A digital camera with a ring flash was fixed in a frame, so that the tongue-lens-distance and illumination were kept constant. A pilot study testing the inter-methods reliability, test-retest reliability, and interrater reliability of 2 observers was conducted: the tongues (1 ×) as well as the tongue images (2 ×) from 76 patients from our ward were assessed with a standardized rating list in a randomized order by both observers, who were mutually blinded. As primary outcome measure we used agreement beyond chance (Cohen's kappa).

Results: The colors of the tongue body / the tongue coating were the main criteria for the authentic reproduction of the tongue in our images. Inter-methods reliability for the color of the tongue body/coating was kappa 0.35 / 0.34 (p < 0.001) for observer 1 and 0.59 / 0.57 (p < 0.001) for observer 2. Test-retest reliability for the color of the tongue body / the tongue coating was kappa 0.53 / 0.48 (p< 0.001) for observer 1 and 0.65 / 0.62 (p < 0.001) for observer 2. Interrater reliability was generally low.

Conclusions: The introduced device represents a first step towards standardization of tongue diagnosis. However, inter-methods as well as test-retest reliability vary between observers and different morphological characteristics.