Tongue Diagnosis Research Articles

Objective diagnosis of tongue in Kampo Medicine

Objective diagnosis of tongue in Kampo Medicine

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application.

Tongue diagnosis is an essential technique of traditional Chinese medicine (TCM) diagnosis, and the need for objectifying tongue images through image processing technology is growing. The present study provides an overview of the progress made in tongue objectification over the past decade and compares segmentation models. Various deep learning models are constructed to verify and compare algorithms using real tongue image sets. The strengths and weaknesses of each model are analyzed. The findings indicate that the U-Net algorithm outperforms other models regarding precision accuracy (PA), recall, and mean intersection over union (MIoU) metrics. However, despite the significant progress in tongue image acquisition and processing, a uniform standard for objectifying tongue diagnosis has yet to be established. To facilitate the widespread application of tongue images captured using mobile devices in tongue diagnosis objectification, further research could address the challenges posed by tongue images captured in complex environments.

A two-stage segmentation of sublingual veins based on compact fully convolutional networks for Traditional Chinese Medicine images.

As one of the key methods of Traditional Chinese Medicine inspection, tongue diagnosis manifests the advantages of simplicity and directness. Sublingual veins can provide essential information about human health. In order to automate tongue diagnosis, sublingual veins segmentation has become one important issue in the field of Chinese medicine medical image processing. At present, the primary methods for sublingual veins segmentation are traditional feature engineering methods and the feature representation methods represented by deep learning. The former, which mainly based on colour space, belongs to unsupervised classification method. The latter, which includes U-Net and other deep neural network models, belongs to supervised classification method. Current feature engineering methods can only capture low dimensional information, which makes it difficult to extract efficient features for sublingual veins. On the other hand, current deep learning methods use down-sampling structures, which manifest weak robustness and low accuracy. So, it is difficult for current segmentation approaches to recognize tiny branches of sublingual veins. To overcome the above limits, this paper proposes a novel two-stage semantic segmentation method for sublingual veins. In the first stage, a fully convolutional network without down-sampling is used to realize the accurate segmentation of the tongue that includes the sublingual veins to be segmented in the next stage. During the tongue segmentation, the proposed networks can effectively reduce the loss of medical images spatial feature information. At the same time, in order to expand the receptive field, the dilated convolution has been introduced to the proposed networks, which can capture multi-scale information of segmentation images. In the second stage, another fully convolutional network has been used to segment the sublingual veins on the base of the results from the first stage. In this model, proper dilated convolutional rates have been selected to avoid gridding issue. In order to keep the quality of the images to be segmented, several particular data pre-processing and post-processing have been used, which includes specular highlight removal, data augmentation, erosion and dilation. Finally, in order to evaluate the performance of the proposed model, segmentation results have been compared with the state-of-the-art methods on the base of the dataset from Shanghai University of Traditional Chinese Medicine. The effectiveness of sublingual veins segmentation has been proved.

Objective evaluation of tongue diagnosis ability using a tongue diagnosis e-learning/e-assessment system based on a standardized tongue image database.

In Kampo medicine, tongue examination is used to diagnose the pathological condition "Sho," but an objective evaluation method for its diagnostic ability has not been established. We constructed a tongue diagnosis electronic learning and evaluation system based on a standardized tongue image database. This study aims to verify the practicality of this assessment system by evaluating the tongue diagnosis ability of Kampo specialists (KSs), medical professionals, and students. In the first study, we analyzed the answer data of 15 KSs in an 80-question tongue diagnosis test that assesses eight aspects of tongue findings and evaluated the (i) test score, (ii) test difficulty and discrimination index, (iii) diagnostic consistency, and (iv) diagnostic match rate between KSs. In the second study, we administered a 20-question common Kampo test and analyzed the answer data of 107 medical professionals and 56 students that assessed the tongue color discrimination ability and evaluated the (v) correct answer rate, (vi) test difficulty, and (vii) factors related to the correct answer rate. In the first study, the average test score was 62.2 ± 10.7 points. Twenty-eight questions were difficult (correct answer rate, <50%), 34 were moderate (50%-85%), and 18 were easy (≥85%). Regarding intrarater reliability, the average diagnostic match rate of five KSs involved in database construction was 0.66 ± 0.08, and as for interrater reliability, the diagnostic match rate between the 15 KSs was 0.52 (95% confidence interval, 0.38-0.65) for Gwet's agreement coefficient 1, and the degree of the match rate was moderate. In the second study, the difficulty level of questions was moderate, with a correct rate of 81.3% for medical professionals and 82.1% for students. The discrimination index was good for medical professionals (0.35) and poor for students (0.06). Among medical professionals, the correct answer group of this question had a significantly higher total score on the Kampo common test than the incorrect answer group (85.3 ± 8.4 points vs. 75.8 ± 11.8 points, p < 0.01). This system can objectively evaluate tongue diagnosis ability and has high practicality. Utilizing this system can be expected to contribute to improving learners' tongue diagnosis ability and standardization of tongue diagnosis.

Discussion on Technical Evaluation of Tongue Diagnosis Equipment of Traditional Chinese Medicine

This study briefly introduces the tongue diagnostic equipment of traditional Chinese medicine. It analyzes and discusses the key points of technical evaluation of tongue diagnostic equipment from the aspects of product name, performance parameters, image processing functions, product use methods, clinical evaluation, etc. It analyzes the safety risks and effectiveness indicators of tongue diagnostic equipment, hoping to bring some help to the gradual standardization of tongue diagnostic equipment and the registration of enterprises.

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.

Tongue crack recognition using segmentation based deep learning

Tongue cracks refer to fissures with different depth and shapes on the tongue’s surface, which can characterize the pathological characteristics of spleen and stomach. Tongue cracks are of great significance to the objective study of tongue diagnosis. However, tongue cracks are small and complex, existing methods are difficult to extract them effectively. In order to achieve more accurate extraction and identification of tongue crack, this paper proposes to apply a deep learning network based on image segmentation (Segmentation-Based Deep-Learning, SBDL) to extract and identify tongue crack. In addition, we have studied the quantitative description of tongue crack features. Firstly, the pre-processed tongue crack samples were amplified by using adding salt and pepper noise, changing the contrast and horizontal mirroring; secondly, the annotation tool Crack-Tongue was used to label tongue crack; thirdly, the tongue crack extraction model was trained by using SBDL; fourthly, the cracks on the tongue surface were detected and located by the segmentation network, and then the output and features of the segmentation network were put into the decision network for the classification of crack tongue images; finally, the tongue crack segmentation and identification results were quantitatively evaluated. The experimental results showed that the tongue crack extraction and recognition results based on SBDL were better than Mask Region-based Convolutional Neural Network (Mask R-CNN), DeeplabV3+, U-Net, UNet++ and Semantic Segmentation with Adversarial Learning (SegAN). This method effectively solved the inaccurate tongue crack extraction caused by the tongue crack’s color being close to the surrounding tongue coating’s color. This method can achieve better tongue crack extraction and recognition results on a small tongue crack data set and provides a new idea for tongue crack recognition, which is of practical value for tongue diagnosis objectification.

A survey of artificial intelligence in tongue image for disease diagnosis and syndrome differentiation.

The rapid development of artificial intelligence technology has gradually extended from the general field to all walks of life, and intelligent tongue diagnosis is the product of a miraculous connection between this new discipline and traditional disciplines. We reviewed the deep learning methods and machine learning applied in tongue image analysis that have been studied in the last 5 years, focusing on tongue image calibration, detection, segmentation, and classification of diseases, syndromes, and symptoms/signs. Introducing technical evolutions or emerging technologies were applied in tongue image analysis; as we have noticed, attention mechanism, multiscale features, and prior knowledge were successfully applied in it, and we emphasized the value of combining deep learning with traditional methods. We also pointed out two major problems concerned with data set construction and the low reliability of performance evaluation that exist in this field based on the basic essence of tongue diagnosis in traditional Chinese medicine. Finally, a perspective on the future of intelligent tongue diagnosis was presented; we believe that the self-supervised method, multimodal information fusion, and the study of tongue pathology will have great research significance.

Simulated Annealing with Deep Learning Based Tongue Image Analysis for Heart Disease Diagnosis

Tongue image analysis is an efficient and non-invasive technique to determine the internal organ condition of a patient in oriental medicine, for example, traditional Chinese medicine (TCM), Japanese traditional herbal medicine, and traditional Korean medicine (TKM). The diagnosis procedure is mainly based on the expert's knowledge depending upon the visual inspection comprising color, substance, coating, form, and motion of the tongue. But conventional tongue diagnosis has limitations since the procedure is inconsistent and subjective. Therefore, computer-aided tongue analyses have a greater potential to present objective and more consistent health assessments. This manuscript introduces a novel Simulated Annealing with Transfer Learning based Tongue Image Analysis for Disease Diagnosis (SADTL-TIADD) model. The presented SADTL-TIADD model initially pre-processes the tongue image to improve the quality. Next, the presented SADTL-TIADD technique employed an EfficientNet-based feature extractor to generate useful feature vectors. In turn, the SA with the ELM model enhances classification efficiency for disease detection and classification. The design of SA-based parameter tuning for heart disease diagnosis shows the novelty of the work. A wide-ranging set of simulations was performed to ensure the improved performance of the SADTL-TIADD algorithm. The experimental outcomes highlighted the superior of the presented SADTL-TIADD system over the compared methods with maximum accuracy of 99.30%.

LAIU-Net: A learning-to-augment incorporated robust U-Net for depressed humans’ tongue segmentation

Computer-aided tongue diagnosis system requires segmentation of the tongue body. The frequent movement of the tongue due to its natural flexibility often causes shape variability in photographs across subjects, which makes segmenting the tongue challenging from non-tongue elements, such as the lips, teeth, and other objects in the background of the tongue. The flexibility of the tongue causes a further challenge in maintaining a similar shape and style when taking photos of many healthy subjects and patients. To address these challenges, we have built a tongue dataset, where the tongue of each subject has been scanned thrice with an interval of less than a second. We have collected 333 tongue images from 111 depressed humans, who have been diagnosed with depression by a psychiatrist. In addition, in this paper, we propose a learning-to-augment incorporated U-Net (LAIU-Net) for the segmentation of the depressed human tongue in photographic images. The best policies for data augmentation were automatically chosen with the proposed LAIU-Net. For this purpose, we corrupted photographic tongue images with the Gaussian, speckle, and Poisson noise. The proposed approach addresses the overfitting problem as well as increases the generalizability of a deep network. We have compared the performance of the proposed LAIU-Net with that of other state-of-the-art U-Net configurations. Our LAIU-Net approach achieved a mean boundary F1 score of 93.1%.

Tongue Diagnosis using CNN for Disease Detection

In this modern lifestyle, technologies are helping us to maintain our finances, our household things, shopping, and so on. In our research work, we have proposed an application that would tell you the disease or infection that you may have with the help of the developing technology. In this pandemic period, we have to be safer and more Responsible. We have to avoid visiting public places as much as possible for us and our society. Our main aim is to reduce death rates which are all caused due to finding the disease at its final stage because of hesitation to visit the hospital during this pandemic or because of our carelessness. We can overcome it by checking for diseases or infections frequently using a mobile app. In this research work, we are planning to develop a mobile application using which we can frequently check for diseases or infections since we always have our mobile phones with us. With this application, we can detect the percentage of chance of disease that the user may have through tongue diagnosis by considering changes in various tongue factors. The basic objective of the research work is to make people know about their body condition at an earlier stage more easily and quickly with their smart mobile. In this report, we have included the literature survey made for this proposed system, existing works, software requirements, the proposed system, etc.

Administration of &lt;italic&gt;Yijung-tang&lt;/italic&gt;, &lt;italic&gt;Pyeongwi-san&lt;/italic&gt;, and &lt;italic&gt;Shihosogan-tang&lt;/italic&gt; for Standardization of Korean Medicine Pattern Identification for Functional Dyspepsia: A Study Protocol of a Randomized, Assessor-blind, 3-Arm, Parallel, Open-label, Multicenter Clinical Trial

Objectives: The purpose of this study is to explore the effectiveness and safety of frequently used clinical herbal medicines (Yijung-tang [Lizhong-tang, LJT], Pyeongwi-san [Pingwei-san, PWS], and Shihosogan-tang [Chaihu Shugan-tang, SST]) in patients with functional dyspepsia (FD) when administered according to herbal medicine and Korean medicine pattern identification. The results of this study will be used to standardize the diagnostic instrument used in Korean medicine and to investigate biomarkers of Korean medicine pattern identification.Methods: This study will be a randomized, assessor-blind, 3-arm, parallel, open-label, multi-center clinical trial. A total of 300 FD participants will be recruited from 3 Korean medical hospitals and assigned to the LJT (n=100), PWS (n=100), and SST (n=100) groups according to FD pattern identification. The patients will take the medication for 8 weeks, 3 times a day, before or between meals. The primary outcome will be total dyspepsia symptom (TDS) and the secondary outcomes will be adequate relief (AR) for dyspepsia, overall treatment effect (OTE), visual analogue scale (VAS), functional dyspepsia-related quality of life (FD-QoL), gastrointestinal symptom score (GIS), and pattern identification questionnaires. For the exploratory outcomes, we will analyze blood and fecal metabolome profiles, microbiota from fecal and saliva samples, single nucleotide polymorphism (SNP), and results of Korean medicine diagnosis device measurements (heart rate variability, and tongue, pulse, and abdominal diagnosis).Conclusions: The results of this study will prove objectivity for Korean medicine pattern identifications, and the effectiveness and safety of herbal medicines for the population with FD.

Relationship between thick or greasy tongue-coating microbiota and tongue diagnosis in patients with primary liver cancer.

Tongue diagnosis is a unique aspect of traditional Chinese medicine for diagnosing diseases before determining proper means of treatment, but it also has the disadvantage of relying on the subjective experience of medical practitioners and lack objective basis. The purpose of this article is to elucidate tongue-coating microbiota and metabolic differences in primary liver cancer (PLC) patients with thick or greasy tongue coatings. Tongue-coating samples were analyzed in 60 PLC patients (30 PLC with thick or greasy tongue-coating patients and 30 PLC with tongue-coating neither thick nor greasy) and 25 healthy controls (HC) using 16S rRNA gene sequencing technology. As compared to healthy individuals, tongue coatings of patients with PLC had elevated levels of Firmicutes and Actinobacteria. The abundance of Fusobacteria, SR1_Absconditabacteria_, and Spirochaete were higher in tongue coatings of healthy controls compared to samples in patients with PLC. In addition to site-specific differences, higher abundances of Fusobacteria and Actinobacteria were observed in thick or greasy tongue-coating patients as compared to non-thick and greasy tongue-coating patients. The inferred metagenomic pathways enriched in the PLC tongue-coating patients were mainly those involved in replication, recombination, and repair of protein. We also identify a tongue-coating microbiome signature to discriminate HC and PLC, including 15 variables on genus level. The prediction performance of the signature showed well in the training and validation cohorts. This research illustrates specific clinical features and bacterial structures in PLC patients with different tongue coatings, which facilitates understanding of the traditional tongue diagnosis.

Tongue diagnosis models in Persian medicine: A Delphi study

IntroductionIn Persian Medicine (PM), diagnosis of physiological characteristics and disease symptoms is achieved by clinical indices. This study aimed to present selected tongue images as the first educational model in this area of PM. MethodsIn this Delphi study, tongue diagnostic indices were extracted from reliable PM sources. Twenty PM experts were interviewed regarding their clinical experience on the subject. Tongue photographs were taken, standardized and printed to the extent of data saturation. In face-to-face interviews with experts, they selected some images and described their characteristics. Images selected by at least 25% of the interviewees were allocated to reach a consensus on gathered opinions. Subsequently, images with highest agreement among experts regarding a specific diagnostic index were presented as an educational model. ResultsTongue color, thickness and color of coating, size, shape, moisture, dryness and fissures are most commonly used by PM physicians. Out of 221 standardized images in the first stage, 54 and then 13 images were selected as typical educational models. ConclusionAs the initial steps of standardization of PM diagnostic indices, selected tongue images can be used for educational and research purposes in PM diagnostics. However, future studies are need to increase the accuracy of diagnostic models.

Tongue diagnosis based on hue-saturation value color space: controlled study of tongue appearance in patients treated with percutaneous coronary intervention for coronary heart disease

ObjectiveTo analyze the characteristics of tongue imaging color parameters in patients treated with percutaneous coronary intervention (PCI) and non-PCI for coronary atherosclerotic heart disease (CHD), and to observe the effects of PCI on the tongue images of patients as a basis for the clinical diagnosis and treatment of patients with CHD. MethodsThis study used a retrospective cross-sectional survey to analyze tongue photographs and medical history information from 204 patients with CHD between November 2018 and July 2020. Tongue images of each subject were obtained using the Z-BOX Series traditional Chinese medicine (TCM) intelligent diagnosis instruments, the SMX System 2.0 was used to transform the image data into parameters in the HSV color space, and finally the parameters of the tongue image between patients in the PCI-treated and non-PCI-treated groups for CHD were analyzed. ResultsAmong the 204 patients, 112 were in the non-PCI treatment group (38 men and 74 women; average age of (68.76 ± 9.49) years), 92 were in the PCI treatment group (66 men and 26 women; average age of (66.02 ± 10.22) years). In the PCI treatment group, the H values of the middle and tip of the tongue and the overall coating of the tongue were lower (P < 0.05), while the V values of the middle, tip, both sides of the tongue, the whole tongue and the overall coating of the tongue were higher (P < 0.05). ConclusionThe color parameters of the tongue image could reflect the physical state of patients treated with PCI, which may provide a basis for the clinical diagnosis and treatment of patients with CHD.

Tongue features of patients with granulomatous lobular mastitis.

Traditional Chinese tongue diagnosis plays an irreplaceable role in disease diagnosis. This study aimed to describe the tongue characteristics of patients with granulomatous lobular mastitis (GLM). Forty GLM patients and 40 non-GLM controls were evaluated using the Traditional Chinese Medicine subjective clinical interpretation and a TDA-1 Tongue Diagnostic and Analysis system. The associations between the image features of the tongue body and coating and the profiling of immune-inflammatory parameters were analyzed. GLM patients were prone to a reddish tongue bodies with thick, white, and greasy coatings. Thick and greasy tongue coating features are risk factors for GLM. GLM patients had higher levels of white blood cells (WBC), platelets, C-reactive protein, interleukin-2, and transforming growth factor-β (TGF-β) than non-GLM controls (P < .05). Also, tongue coating contrast and entropy values were significantly correlated with WBC or TGF-β levels in GLM patients (r < -0.310 and P < .05). We demonstrated that the hot evil and phlegm-dampness constitutions are the main characteristics of GLM. This might provide a reference for GLM diagnosis.

Machine Learning-Based Technique for the Severity Classification of Sublingual Varices according to Traditional Chinese Medicine.

Tongue diagnosis, a noninvasive examination, is an essential step for syndrome differentiation and treatment in traditional Chinese medicine (TCM). Sublingual vein (SV) is examined to determine the presence of blood stasis and blood stasis syndrome. Many studies have shown that the degree of SV stasis positively correlates with disease severity. However, the diagnoses of SV examination are often subjective because they are influenced by factors such as physicians' experience and color perception, resulting in different interpretations. Therefore, objective and scientific diagnostic approaches are required to determine the severity of sublingual varices. This study aims at developing a computer-assisted system based on machine learning (ML) techniques for diagnosing the severity of sublingual varicose veins. We conducted a comparative study of the performance of several supervised ML models, including the support vendor machine, K-neighbor, decision tree, linear regression, and Ridge classifier and their variants. The main task was to differentiate sublingual varices into mild and severe by using images of patients' SVs. To improve diagnostic accuracy and to accelerate the training process, we proposed using two model reduction techniques, namely, the principal component analysis in conjunction with the slice inverse regression and the convolution neural network (CNN), to extract valuable features during the preprocessing of data. Our results showed that these two extraction methods can reduce the training time for the ML methods, and the Ridge-CNN method can achieve an accuracy rate as high as 87.5%, which is similar to that of experienced TCM physicians. This computer-aided tool can be used for reference clinical diagnosis. Furthermore, it can be employed by junior physicians to learn and to use in clinical settings.

Study on syndrome rules of stagnated heat in liver and stomach of non-erosive reflux disease based on microecology of tongue coating.

At present, the pathogenesis of non-erosive reflux disease (NERD) is still unclear, and proton pump inhibitors are the main treatment drug. However, the effect is limited. Traditional Chinese medicine treatment of NERD has advantages. Stagnated heat in liver and stomach syndrome is the most important traditional Chinese medicine syndrome type of this disease. Tongue diagnosis is an important basis for the diagnosis of stagnated heat in liver and stomach syndrome. The microecology of tongue coating suggests the occurrence and development of disease. The purpose of this study aims to clarify the regular changes of tongue coating microecology in stagnated heat in liver and stomach syndrome of NERD. This is a cross-sectional clinical trial. This study is divided into NERD stagnated heat in liver and stomach syndrome group, qi stagnation, and phlegm obstruction syndrome control group and normal control group, with 20 cases in each group. Tongue coating samples will be collected from 3 groups, and 16SrRNA gene sequencing technology will be used to detect the genome of tongue coating flora in patients with NERD with stagnated heat in liver and stomach syndrome, control group with qi stagnation and phlegm obstruction syndrome and normal control group. The main outcome measures are the distribution, diversity, and richness of the tongue flora in patients and healthy controls. The results of this study will clarify the internal relationship between NERD stagnated heat in liver and stomach syndrome and the microecological changes in tongue coating.

Black‐brown hairy tongue: An unusual adverse effect of linezolid

AbstractLinezolid‐induced black hairy tongue (BHT) is an uncommon side effect. We present a case of linezolid induced BHT and evaluate relevant literature from the home and worldwide. Its goal is to offer a secure and appropriate foundation for clinical drug usage. A 23‐year‐old male with recurrent boil infection erroneously reported by gram‐positive methicillin‐resistant Staphylococcus aureus (MRSA) by pus culture, which was misdiagnosed by a small clinic (diagnostic center) in a rural region of Bangladesh where he had previously been treated for that reoccurring boil infection. The patient was actually suffering from a different pus‐forming bacteria (Klebsiella), a gram‐negative bacterium that was just found after careful C/S report. Linezolid was administered to the patient 600 mg q12h orally in tablet after discharge from that rural clinic. On the 14th day following oral treatment, the patient presented to the hospital with BHT but no other atypical taste complaints. However, all symptoms were tolerable, and he finished the linezolid treatment period. After withdrawing from the medicine, all tongue symptoms vanished. Following a thorough study, it was determined that this patient had BHT brought on by drug linezolid. Long‐term linezolid usage is associated with neuropathies and bone marrow suppression. There is, however, little information about an uncommon adverse effect, black hairy tongue. We present here a case of linezolid induced BHT in a patient who was misdiagnosed. Also covered are the etiology, pathophysiology, diagnosis, and management of black hairy tongue.

A novel tongue feature extraction method on mobile devices

In Traditional Chinese Medicine (TCM), tongue diagnosis is essential for symptom differentiation and treatment selection. Compared with traditional tongue diagnostic instruments, deploying a tongue diagnosis system on mobile devices is more convenient to monitor general health and facilitates the development of telemedicine. However, limited by both the quality and quantity of tongue images taken by mobile devices, extracting tongue features of the images maintains a great challenge. In this paper, we present a tongue feature extraction method on mobile devices, containing a high-accuracy tongue segmentation method based on Moment Invariants with Data Augmentation (DAMI) and an efficient and lightweight feature classification model with an attention mechanism. Meanwhile, we construct a novel tongue image dataset from mobile devices for extracting tongue features, significantly, first including sublingual images which are beneficial to extracting sublingual vein features. Extensive experiments on two datasets demonstrate the effectiveness and robustness of our method. Furthermore, our method greatly reduces the computing and storage demands compared with other current methods, providing a good prerequisite for deployment on mobile devices. Finally, to demonstrate the potential application of our proposed method, we develop a TCM intelligence tongue diagnosis application, which can be accessed through the WeChat Mini Program or web version, exhibiting its great potential in clinical diagnosis and health monitoring.