Tongue diagnosis is an important procedure in traditional Korean medicine (TKM). In particular, tongue coating thickness (TCT) is deemed to show the progression of the disease. However, conventional tongue diagnosis has limitations because of various external factors. Therefore, it is necessary to investigate the availability of tongue diagnosis system (TDS) in the assessment of TCT. This study has been designed as a prospective clinical trial involving 60 patients with functional dyspepsia. Tongue images will be obtained by TDS twice with a 30 min interval. The system will measure the percentage of TCT and classify it as either no coating, thin coating, or thick coating according to the existing diagnostic criteria. After finishing the collection of 60 patients' tongue images, TCT on the images will be simultaneously evaluated by the conventional method to establish the gold standard for assessing TCT by 5 well-trained clinicians. The evaluation will be repeated by the same clinicians after 2 weeks, but the order of the images will be changed. This trial is expected to provide clinical evidence for the availability of TDS as a diagnostic tool and to contribute to the standardization of the diagnosis system used in TKM. This trial is registered with ClinicalTrials.gov
Tongue diagnosis is an important diagnostic procedure that is used in traditional Korean medicine (TKM) to examine the physiological function and pathological changes of the internal organs. It involves the examination of the body of the tongue (its color, shape, moisture, and movement) and the coating on the tongue (its color, thickness, distribution, and characteristics at the root) [
To overcome the above limitations, a number of tongue diagnosis systems (TDSs) have been developed in recent years [
Functional dyspepsia (FD) is chronic or recurrent dyspepsia without evidence of organic disease [
For the objective and standardized tongue diagnosis system in TKM, it is necessary not only to develop an objective diagnostic tool but also to propose a new study protocol for clinical trial. We aimed to investigate the agreement between the results provided by TDS and those obtained by the conventional method in the assessment of TCT, which constitutes a part of tongue diagnosis in TKM, for evaluating on the availability of TDS in practice.
This study has been designed as a prospective clinical trial on the availability of TDS as a new diagnostic tool. Because there are no data available in the literature that would indicate the effective sample size to investigate the availability of TDS, we estimated the approximate sample size based on the number of outpatients with FD presenting at the Korean Medicine Hospital of the Kyung Hee University in Seoul, Korea.
The study will include a total of 60 patients aged 20 years and older who meet the Rome III criteria for FD [
The exclusion criteria are as follows: pregnancy, mental disorders, severe systemic organ diseases such as cancer, infectious or other serious consumptive diseases, and geographic tongue. Patients unable to open their mouths or protrude their tongues; those unable to read, write, hear, or see; and those unable to sign a written informed consent form will also be excluded.
The study will be conducted in accordance with the standards of the International Committee on Harmonization of Good Clinical Practice and the revised version of the Declaration of Helsinki. The protocol of the trial has been approved by the institutional review board of the Korean Medicine Hospital of the Kyung Hee University (IRB no. KOMCIRB-2013-01). A written informed consent will be obtained from all patients before enrollment.
All patients will be required to visit the hospital twice during the study period. At the first visit, the written informed consent will be obtained, and patients will be screened for eligibility criteria. All eligible patients will be then instructed to avoid mouth rinsing as well as food and liquid intake for 4 hours before the examination with the TDS at the next visit.
Tongue images will be obtained with the TDS under constant conditions at the second visit. Then, the percentage of tongue coating will be measured using a TDS analysis software, and TCT will be classified with the existing diagnostic criteria [
After we complete the collection of 60 patients’ tongue images, the images will be simultaneously evaluated by 5 clinicians with 5 years or more of clinical experience to establish the gold standard for the assessment of TCT. The evaluation will be repeated after 2 weeks to analyze intra- and interrater reliability of the clinicians. Finally, we will evaluate the agreement between the results provided by TDS and those obtained by clinicians. The flow of the trial is shown in Figure
Flow chart of the trial. *TDS: tongue diagnosis system.
TDS will measure the percentage of tongue coating and classify TCT as either no coating, thin coating, or thick coating according to the existing diagnostic criteria [
To evaluate the performance of TDS, we will assess its reproducibility by comparing the 2 measurements of TCT in each patient and analyze the correlation between the percentage of coating measured with the TDS and the weight of coating obtained by tongue scraping [
The TDS consists of image acquisition system, LED illuminator, case, and analysis software. It is equipped with a vision camera (HVR-2130CPA, Hyvision System, Korea) and a mount lens (H2Z0414C-MP, Hyvision System, Korea). The camera can take color images of SXGA resolution (
Components of a tongue diagnosis system (TDS).
An analysis software program of TDS was developed to analyze the percentage of tongue coating from the captured tongue images. To calculate the percentage of tongue coating, we first extract the tongue area from the captured tongue image. TDS generates 17 nodes on the tongue image on the monitor, and the operator moves the nodes on the boundary line of the tongue with a mouse. Then, the remaining area excluding the selected tongue area is removed. TDS converts the chromaticity coordinates of the extracted tongue image in the RGB color space to those in the CIE-Lab color space. The area of the tongue body is clearly more reddish than that of the coating. The positive a* value in the CIE-Lab color space is indicated by red/magenta, while the negative value by green. Therefore, tongue coating area is extracted from the entire tongue based on the difference in the a* value. The binarization threshold point of the a* value for the extraction was set through the tongue images acquired from the previous study [
Schematic diagram for calculating tongue coating percentage by extracting tongue coating area.
Patients press their face against the interface of the TDS, open their mouth wide, and protrude their tongue for a second. An operator who is well trained in the use of TDS makes sure that the patient’s tongue does not bend or lean to one side by watching the tongue image on the monitor and takes a picture of the tongue when ready. TDS allows to measure tongue coating percentage and classify TCT into 1 of 3 categories—no coating, thin coating, or thick coating—according to the diagnostic criteria developed by Kim et al. [
To assess the reproducibility of TDS, tongue images will be taken twice with a 30 min interval. Considering inevitable variations in length, angle, and shape of the protruding tongue in taking images, it will be necessary to have an appropriate interval between the two examinations to more accurately investigate the reproducibility of TDS. A 30 min interval will be able to prevent patients from intentionally protruding their tongues as they had done in the first examination. Then, we will compare TCT as shown on two different images for each patient.
After two examinations of TDS, we will measure the weight of the tongue coating with a method described by Yaegaki and Sanada [
In the conventional method of tongue diagnosis, including the evaluation of tongue coating, a practitioner directly examines the tongue from the tip to the root. TCT is described as either no, thin, or thick coating. No coating literally means the absence of tongue coating; thin coating is diagnosed when the body of the tongue is barely visible; and thick coating when the body of the tongue cannot be seen at all [
To establish the gold standard for the assessment of TCT, 5 clinicians with 5 or more years of clinical experience and working at the Korean Medicine Hospital of the Kyung Hee University will be included in the trial to assess the images obtained by TDS. Before participation in the study, they will augment their knowledge about conventional methods of TCT assessment and be well trained in TCT differential criteria to obtain a reliable gold standard. The images will be sequentially presented to the clinicians using a picture-viewer software and an LCD monitor (DN-50PZ66,
The baseline characteristics of the patients will be presented as numbers and percentages for categorical variables and as mean ± standard deviation for continuous variables. Cohen’s kappa
Tongue diagnosis is an important procedure in TKM, but conventional methods are limited by various external factors [
The study was designed in 2013. The first participant was enrolled in February, 2013. The recruitment of participants is ongoing. In addition, we have already finished the clinical trial registration (ClinicalTrials.gov
The authors declare that they have no competing interests.
This study was supported by a Grant of the Oriental Medicine Advanced Technology R&D Project, Ministry for Health, Welfare and Family Affairs, Republic of Korea (B100030).