The association between objective tongue color and endoscopic findings: results from the Kyushu and Okinawa population study (KOPS)

Cross-population tongue image features and tongue coating microbiome changes in the evolution of colorectal cancer

Introduction

Tongue diagnosis, a cornerstone of Traditional Chinese Medicine (TCM), relies significantly on the assessment of tongue coating, which is used to evaluate Zang-fu organ functions, qi and blood dynamics, and the influence of pathogenic factors. This diagnostic method is integral to disease diagnosis and treatment in TCM. Recent research suggests a strong correlation between the characteristics of tongue coating and its microbial composition. These microbial variations may influence the formation and changes in tongue coating and are potentially linked to the progression of specific diseases. However, comprehensive research on the association between tongue coating, its microorganisms, and colorectal cancer (CRC) is limited. Notably, the quantitative aspects of tongue diagnosis and the microbial diversity in tongue coatings across different stages of colorectal cancer (from healthy individuals to colorectal adenoma (CRA) and CRC patients) are yet to be fully elucidated. By studying the cross-population characteristics of tongue image and tongue coating microorganisms during the evolution of colorectal cancer, the differences of tongue image characteristics and tongue coating microorganisms among different populations were further evaluated, providing references for early screening, diagnosis and treatment of colorectal cancer.

Methods

The tongue image features of the subjects were collected by DS01-B tongue surface information collection system, mainly including tongue quality and tongue coating, and the tongue image was quantitatively analyzed by color space Lab value. The microbial characteristics of tongue coating were detected by high-throughput sequencing (16SrRNA amplicon sequencing). All subjects came from the patients in the Sixth Affiliated Hospital of Sun Yat-sen University and recruited volunteers (divided into health group, CRA group and CRC group), and obtained the ethical approval of the Sixth Affiliated Hospital of Sun Yat-sen University (ethical batch number: 2021ZSLYEC-328).

Results

A total of 377 subjects were recruited in this study, including 56 healthy subjects, 65 colorectal adenomas and 256 colorectal cancer patients. The results showed that: in terms of texture of fur, the "thick fur" was a significant statistical difference (p < 0.05) in the 3 groups. In addition, there was also a statistical difference in "greasy fur" and "peeled fur" among the 3 groups (p < 0.05). Lab quantitative analysis of tongue color and fur color: The results showed that the L value of tongue color in healthy group was significantly different from that in CRA group and CRC group (p < 0.01), but there was no significant difference between CRA group and CRC group (p > 0.05). Tongue coating microorganisms, there was no significant difference in the richness and diversity of the three groups of subjects (p > 0.05). There were 296 species in the three groups, accounting for 44.65%, and the species in colorectal cancer population was the most, reaching 502. From the differences in community composition among the three groups, it was found that there were certain differences in bacterial community composition between healthy people, CRA and CRC, and the differences became more and more obvious with the development of the disease.

Conclusion

This study revealed the specific cross-population tongue image characteristics and the specificity of tongue coating microorganisms in the evolution of CRC, providing new research ideas for early screening, early diagnosis, mechanism exploration, prevention and treatment of colorectal cancer.

The association between oral microbiome and gastric precancerous lesions

Gastric precancerous lesions are thought to be precursors in the occurrence and development of gastric cancer through Correa's cascade. Recent studies have investigated the association between the oral microbiome and gastric precancerous lesions. However, there has yet to be a comprehensive synthesis review of the existing literature on the relationship between oral microbiome and gastric precancerous lesions. A systematic review was conducted to characterize the literature on the association between oral microbiome and gastric precancerous lesions. The studies show that oral microbiome is dynamic in individuals with gastric precancerous lesions. Oral-derived microorganisms were colonized in the gastric precancerous lesions. Interactions between oral and gastric microbiomes affect the response of the host immunity. The abnormal proliferation of oral-associated microorganisms may be linked to the reduction of gastric acid. The present review supports the potential association between oral microbiome and gastric precancerous lesions. However, the interactions are complex and multifaceted, which require further investigation.

Development of an artificial intelligent model for pre-endoscopic screening of precancerous lesions in gastric cancer

BACKGROUND

Given the high cost of endoscopy in gastric cancer (GC) screening, there is an urgent need to explore cost-effective methods for the large-scale prediction of precancerous lesions of gastric cancer (PLGC). We aim to construct a hierarchical artificial intelligence-based multimodal non-invasive method for pre-endoscopic risk screening, to provide tailored recommendations for endoscopy.

METHODS

From December 2022 to December 2023, a large-scale screening study was conducted in Fujian, China. Based on traditional Chinese medicine theory, we simultaneously collected tongue images and inquiry information from 1034 participants, considering the potential of these data for PLGC screening. Then, we introduced inquiry information for the first time, forming a multimodality artificial intelligence model to integrate tongue images and inquiry information for pre-endoscopic screening. Moreover, we validated this approach in another independent external validation cohort, comprising 143 participants from the China-Japan Friendship Hospital.

RESULTS

A multimodality artificial intelligence-assisted pre-endoscopic screening model based on tongue images and inquiry information (AITonguequiry) was constructed, adopting a hierarchical prediction strategy, achieving tailored endoscopic recommendations. Validation analysis revealed that the area under the curve (AUC) values of AITonguequiry were 0.74 for overall PLGC (95% confidence interval (CI) 0.71-0.76, p < 0.05) and 0.82 for high-risk PLGC (95% CI 0.82-0.83, p < 0.05), which were significantly and robustly better than those of the independent use of either tongue images or inquiry information alone. In addition, AITonguequiry has superior performance compared to existing PLGC screening methodologies, with the AUC value enhancing 45% in terms of PLGC screening (0.74 vs. 0.51, p < 0.05) and 52% in terms of high-risk PLGC screening (0.82 vs. 0.54, p < 0.05). In the independent external verification, the AUC values were 0.69 for PLGC and 0.76 for high-risk PLGC.

CONCLUSION

Our AITonguequiry artificial intelligence model, for the first time, incorporates inquiry information and tongue images, leading to a higher precision and finer-grained pre-endoscopic screening of PLGC. This enhances patient screening efficiency and alleviates patient burden.

Effects of Sitting and Supine Positions on Tongue Color as Measured by Tongue Image Analyzing System and Its Relation to Biometric Information

Tongue diagnosis is one of the important diagnostic methods in Kampo (traditional Japanese) medicine, in which the color and shape of the tongue are used to determine the patient's constitution and systemic symptoms. Tongue diagnosis is performed with the patient in the sitting or supine positions; however, the differences in tongue color in these two different positions have not been analyzed. We developed tongue image analyzing system (TIAS), which can quantify tongue color by capturing tongue images in the sitting and supine positions. We analyzed the effects on tongue color in two different body positions. Tongue color was quantified as L∗a∗b∗ from tongue images of 18 patients in two different body positions by taking images with TIAS. The CIEDE 2000 color difference equation (ΔE00) was used to assess the difference in tongue color in two different body positions. Correlations were also determined between ΔE00, physical characteristics, and laboratory test values. The mean and median ΔE00 for 18 patients were 2.85 and 2.34, respectively. Of these patients, 77.8% had a ΔE00 < 4.1. A weak positive correlation was obtained between ΔE00 and systolic blood pressure and fasting plasma glucose. Approximately 80% of patients' tongue color did not change between the sitting and supine positions. This indicates that the diagnostic results of tongue color are trustworthy even if medical professionals perform tongue diagnosis in two different body positions.

Risk factors and 26-years worldwide prevalence of endoscopic erosive esophagitis from 1997 to 2022: a meta-analysis

Erosive esophagitis (EE) is the part of gastroesophageal reflux disease (GERD) spectrum and may progress to esophageal adenocarcinoma. Due to its progressivity and unclear prevalence, we aim to identify the factors contributing in EE to decide the need for further examination. We performed a PRISMA 2020-based systematic search through PubMed and other resources up to June 2, 2022. Study quality was assessed using the Newcastle-Ottawa Scale (NOS). The odds ratio (OR) of each factor and worldwide prevalence of EE were measured. There are 114 observational studies included with a total of 759,100 participants. Out of 29 factors, the significant risk factors are age ≥ 60 y.o. (OR 2.03 [1.81-2.28]), White/Caucasian (OR 1.67 [1.40-1.99]), unmarried (OR 1.08 [1.03-1.14]), having GERD ≥ 5 years (OR 1.27 [1.14-1.42]), general obesity (OR 1.78 [1.61-1.98]), central obesity (OR 1.29 [1.18-1.42]), diabetes mellitus (DM) (OR 1.24 [1.17-1.32]), hypertension (OR 1.16 [1.09-1.23]), dyslipidemia (OR 1.15 [1.06-1.24]), hypertriglyceridemia (OR 1.42 [1.29-1.57]), hiatal hernia (HH) (OR 4.07 [3.21-5.17]), and non-alcoholic fatty liver disease (NAFLD) (OR 1.26 [1.18-1.34]). However, H. pylori infection (OR 0.56 [0.48-0.66]) and atrophic gastritis (OR 0.51 [0.31-0.86]) are protective towards EE. This study demonstrates that age, ethnicity, unmarried, long-term GERD, metabolic diseases, HH, and NAFLD act as risk factors for EE, whereas H. pylori infection and atrophic gastritis act as protective factors. These findings may enable a better understanding of EE and increase greater awareness to address its growing burden.

Construction of Tongue Image-Based Machine Learning Model for Screening Patients with Gastric Precancerous Lesions

Screening patients with precancerous lesions of gastric cancer (PLGC) is important for gastric cancer prevention. The accuracy and convenience of PLGC screening could be improved with the use of machine learning methodologies to uncover and integrate valuable characteristics of noninvasive medical images related to PLGC. In this study, we therefore focused on tongue images and for the first time constructed a tongue image-based PLGC screening deep learning model (AITongue). The AITongue model uncovered potential associations between tongue image characteristics and PLGC, and integrated canonical risk factors, including age, sex, and Hp infection. Five-fold cross validation analysis on an independent cohort of 1995 patients revealed the AITongue model could screen PLGC individuals with an AUC of 0.75, 10.3% higher than that of the model with only including canonical risk factors. Of note, we investigated the value of the AITongue model in predicting PLGC risk by establishing a prospective PLGC follow-up cohort, reaching an AUC of 0.71. In addition, we developed a smartphone-based app screening system to enhance the application convenience of the AITongue model in the natural population from high-risk areas of gastric cancer in China. Collectively, our study has demonstrated the value of tongue image characteristics in PLGC screening and risk prediction.

Reliability of non-contact tongue diagnosis for Sjögren's syndrome using machine learning method

Sjögren's syndrome (SS) is an autoimmune disease characterized by dry mouth. The cause of SS is unknown, and its diverse symptoms make diagnosis difficult. The Saxon test, an intraoral examination, is used as the primary diagnostic method for SS, however, the risk of salivary infection is problematic. Therefore, we investigate the possibility of diagnosing SS by non-contact and imaging observation of the tongue surface. In this study, we obtained tongue photographs of 60 patients at the Tsurumi University School of Dentistry outpatient clinic to clarify the relationship between the features of the tongue and SS. We divided the tongue into four regions, and the color of each region was transformed into CIE1976L*a*b* space and statistically analyzed. To clarify experimentally the possibility of SS diagnosis using tongue color, we employed three machine-learning models: logistic regression, support vector machine, and random forest. In addition, we constructed diagnostic prediction models based on the Bagging and Stacking methods combined with three machine-learning models for comparative evaluation. This analysis used dimensionality compression by principal component analysis to eliminate redundancy in tongue color information. We found a significant difference between the a* value of the rear part of the tongue and the b* value of the middle part of the tongue in SS and non-SS patients. In addition to the principal component scores of tongue color, the support vector machine was trained using age, and achieved high accuracy (71.3%) and specificity (78.1%). The results indicate that the prediction of SS diagnosis by tongue color reaches a level comparable to machine learning models trained using the Saxon test. This is the first study using machine learning to predict SS diagnosis by non-contact tongue observation. Our proposed method can potentially support early SS detection simply and conveniently, eliminating the risk of infection at diagnosis, and it should be validated and optimized in clinical practice.

Assessment of the Possible Correlation between the Presence of Helicobacter Pylori Infection and Hairy Tongue Lesion in a Group of Patients in Syria: A Cross-Sectional and Pilot Study

BACKGROUND

This study aimed to evaluate the correlation between the presence of hairy tongue and H. pylori infection in patients referring to their blood test based on the serum levels of anti-H pylori IgG antibodies.

METHODS

This cross-sectional study was conducted in the Department of Oral Medicine, University of Damascus Dental School, between February 2021 and January 2022. The sample size of 40 patients (23 males, 17 females), whose ages ranged from 20-79 years with a mean age of 41.5 ± 12 years, was calculated using the G*power 3.1.3, with a statistical power of 80% and a significance level of 0.05. The hairy tongue index was assessed by a visual method based on observing the dorsum tongue appearance. Then, a blood test was performed to detect the presence of H. pylori by Immulite 2000 XPi. Statistical analysis was performed using SPSS software 22.0, Chi-square.

RESULTS

The prevalence of hairy tongue was higher among males (75%) as compared to females (25%) and was found to be statistically significant (p = 0.026). The hairy tongue lesions were found to be least in the 20-39 age group and most prevalent in the 40-59 age group, without statistically significant correlation. H. pylori infection was detected positive in 70% and negative in 30% of hairy tongue patients, compared to the control group, where the rates were 15% and 85%, respectively, with a statistically significant correlation between infection with H. pylori and hairy tongue (p = 0.001).

CONCLUSION

Our results strongly suggest that the hairy tongue might be considered an indicator of H. pylori infection.

A Framework to Predict Gastric Cancer Based on Tongue Features and Deep Learning

Gastric cancer has become a global health issue, severely disrupting daily life. Early detection in gastric cancer patients and immediate treatment contribute significantly to the protection of human health. However, routine gastric cancer examinations carry the risk of complications and are time-consuming. We proposed a framework to predict gastric cancer non-invasively and conveniently. A total of 703 tongue images were acquired using a bespoke tongue image capture instrument, then a dataset containing subjects with and without gastric cancer was created. As the images acquired by this instrument contain non-tongue areas, the Deeplabv3+ network was applied for tongue segmentation to reduce the interference in feature extraction. Nine tongue features were extracted, relationships between tongue features and gastric cancer were explored by using statistical methods and deep learning, finally a prediction framework for gastric cancer was designed. The experimental results showed that the proposed framework had a strong detection ability, with an accuracy of 93.6%. The gastric cancer prediction framework created by combining statistical methods and deep learning proposes a scheme for exploring the relationships between gastric cancer and tongue features. This framework contributes to the effective early diagnosis of patients with gastric cancer.

Oral-Gut Microbiome Analysis in Patients With Metabolic-Associated Fatty Liver Disease Having Different Tongue Image Feature

Objective

The objective of this study was to identify the biological correlation between the tongue coating color and oral and gut micro-characteristics in metabolic-associated fatty liver disease (MAFLD) patients.

Method

The characteristics of the tongue coating were examined using an automatic tongue diagnosis system. Tongue coating and stool samples were collected from 38 MAFLD patients, and 16S rDNA full-length assembly sequencing technology (16S-FAST) was used for bioinformatic analysis.

Results

Twenty-two and 16 subjects were included in two distinct clusters according to the white/yellow color of the tongue coating, which was assessed by the L*a*b* values of the image. Upon analyzing the microorganisms in the tongue coating, 66 and 62 pathognomonic bacterial genera were found in the White and Yellow Coating Groups, respectively. The abundance of Stomatobaculumis positively correlated with the a* values of the tongue coating in the White Coating Group, while Fusobacterium, Leptotrichia, and Tannerella abundance was significantly correlated with the b* values in the Yellow Coating Group. Function prediction mainly showed the involvement of protein families related to BRITE hierarchies and metabolism. The MHR (MONO%/high-density lipoprotein cholesterol) of the Yellow Coating Group was higher than that of the White Coating Group.

Conclusion

In MAFLD patients, lower a* values and higher b* values are indicators of a yellow tongue coating. There were also significant differences in the flora of different tongue coatings, with corresponding changes in the intestinal flora, indicating a correlation between carbohydrate metabolism disorders and inflammation in the oral microbiome.

Combination Image Analysis of Tongue Color and Sublingual Vein Improves the Diagnostic Accuracy of Oketsu (Blood Stasis) in Kampo Medicine

Aim

In tongue diagnosis, a dark purple tongue and enlarged sublingual vein are important findings of Oketsu (blood stasis). However, the association between the tongue color and the sublingual vein has not been reported. This study investigated the association between the tongue color values and the sublingual vein width using tongue image analyzing system (TIAS) for the objective assessment of blood stasis.

Methods

A total of 38 patients (age 68.7 ± 11.3 years, 14 men and 24 women) who visited the Department of Kampo Medicine at Chiba University Hospital were included. Physical findings, blood test results, blood stasis score from medical records, and tongue images obtained with TIAS were analyzed. The patients were classified into two groups: patients with a sublingual vein width of ≤2.5 mm (20 patients) and those with a width of >2.5 mm (18 patients). The physical findings and the blood test results of the two groups were analyzed by Wilcoxon's rank-sum test or χ²-test, whereas logistic regression analysis was used to determine the association between the tongue color values and sublingual vein width. Receiver operating characteristic (ROC) analysis was used to differentiate blood stasis.

Results

The color values significantly related to the sublingual vein width (mm) were the P1-L^* and P4-L^* (darkness of the tongue edge and tongue apex) and the P1-b^* and P2-b^* (blueness of the tongue edge and tongue posterior). The area under the curve was greater for the combination of the tongue color values and the sublingual vein width than that for either of them.

Conclusion

This study demonstrated an objective evaluation of blood stasis in the tongue of patients with dark-blue discoloration and an enlarged sublingual vein. In addition, the combination of the tongue color and the sublingual vein is expected to facilitate a more reliable diagnosis of blood stasis.

GASTRO-CADx: a three stages framework for diagnosing gastrointestinal diseases

Gastrointestinal (GI) diseases are common illnesses that affect the GI tract. Diagnosing these GI diseases is quite expensive, complicated, and challenging. A computer-aided diagnosis (CADx) system based on deep learning (DL) techniques could considerably lower the examination cost processes and increase the speed and quality of diagnosis. Therefore, this article proposes a CADx system called Gastro-CADx to classify several GI diseases using DL techniques. Gastro-CADx involves three progressive stages. Initially, four different CNNs are used as feature extractors to extract spatial features. Most of the related work based on DL approaches extracted spatial features only. However, in the following phase of Gastro-CADx, features extracted in the first stage are applied to the discrete wavelet transform (DWT) and the discrete cosine transform (DCT). DCT and DWT are used to extract temporal-frequency and spatial-frequency features. Additionally, a feature reduction procedure is performed in this stage. Finally, in the third stage of the Gastro-CADx, several combinations of features are fused in a concatenated manner to inspect the effect of feature combination on the output results of the CADx and select the best-fused feature set. Two datasets referred to as Dataset I and II are utilized to evaluate the performance of Gastro-CADx. Results indicated that Gastro-CADx has achieved an accuracy of 97.3% and 99.7% for Dataset I and II respectively. The results were compared with recent related works. The comparison showed that the proposed approach is capable of classifying GI diseases with higher accuracy compared to other work. Thus, it can be used to reduce medical complications, death-rates, in addition to the cost of treatment. It can also help gastroenterologists in producing more accurate diagnosis while lowering inspection time.

Tongue diagnosis indices for gastroesophageal reflux disease: A cross-sectional, case-controlled observational study

Traditional Chinese medicine tongue diagnosis can mirror the status of the internal organ, but evidence is lacking regarding the accuracy of tongue diagnosis to gastroesophageal reflux disease (GERD). This study was to investigate the association between GERD and tongue manifestation, and whether tongue imaging could be initial diagnosis of GERD noninvasively.We conducted a cross-sectional, case-controlled observational study at Kaohsiung Chang Gung Memorial Hospital in Taiwan from January 2016 to September 2017. Participants aged over 20 years old with GERD were enrolled and control group without GERD were matched by sex. Tongue imaging were acquired with automatic tongue diagnosis system, then followed by endoscope examination. Nine tongue features were extracted, and a receiver operating characteristic (ROC) curve, analysis of variance, and logistic regression were used.Each group enrolled 67 participants. We found that the saliva amount (P = .009) and thickness of the tongue's fur (P = .036), especially that in the spleen-stomach area (%) (P = .029), were significantly greater in patients with GERD than in those without. The areas under the ROC curve of the amount of saliva and tongue fur in the spleen-stomach area (%) were 0.606 ± 0.049 and 0.615 ± 0.050, respectively. Additionally, as the value of the amount of saliva and tongue fur in the spleen-stomach area (%) increased, the risk of GERD rose by 3.621 and 1.019 times, respectively. The tongue fur in the spleen-stomach area (%) related to severity of GERD from grade 0 to greater than grade B were 51.67 ± 18.72, 58.10 ± 24.60, and 67.29 ± 24.84, respectively.The amount of saliva and tongue fur in the spleen-stomach area (%) might predict the risk and severity of GERD and might be noninvasive indicators of GERD. Further large-scale, multi-center, randomized investigations are needed to confirm the results.Trial registration: NCT03258216, registered August 23, 2017.

Bioelectrical Impedance Analysis (BIA) of the association of the Japanese Kampo concept "Suidoku" (fluid disturbance) and the body composition of women

Background

In Japanese Kampo medical practice, suidoku (fluid disturbance) is one of the most important concepts for selecting the proper medication. Suidoku is an excessive or uneven distribution of fluid that is indicated by splashing sounds and pitting edema. However, few objective reports about suidoku have been published. Bioelectrical impedance analysis (BIA) uses resistance values obtained from weak electrical currents to estimate body composition, including intracellular and extracellular water and muscle and fat mass. In this study, we used BIA to search for objective factors that can discriminate the various types of suidoku.

Methods

Two hundred twenty-nine patients who visited the Kampo Medicine Clinic of Kyushu University Hospital from June 2010 to August 2015 were divided into non-suidoku (n = 180, 80 male and 100 female), splashing sound (n = 32, 8 male and 24 female) and edema groups (n = 17, 5 male and 12 female). Body composition values were taken from the electronic medical records of InBody730 (a vertical, segmental, multi-frequency analyzer by InBody, Tokyo Japan) testing done at the initial visit. Various parameters of the body composition values of female in the non-suidoku and suidoku groups (splashing sound and edema groups) were compared: there were too few male patients to provide significance.

Results

The age and body weight were significantly lower in the splashing sound group than in the non-suidoku group (p < 0.05). In contrast, the body weight of the edema group was significantly heavier than that of the non-suidoku group (p < 0.05). In ROC analysis, the percent Body Fat ≤ 27.8 %, Muscle Mass Index of the Trunk ≤ 6.5 kg/m², VFA (Visceral fat area) ≤ 5.4 and BMI ≤ 19.2 kg/m² were associated with splashing sound, and Muscle Mass Index of Legs ≥ 4.8 kg/m² and BMI ≥ 21.4 kg/m² were associated with edema.

Conclusion

Our data suggest that the use of this type of BIA to estimate body composition would be a useful tool for the diagnosis of suidoku for women.