A tongue features fusion approach to predicting prediabetes and diabetes with machine learning

Association between color value of tongue and T2DM based on dose-response analyses using restricted cubic splines in China: A cross-sectional study

This study aimed to explore the relationship between international commission on illumination (CIE) L*a*b* color value of tongue and type 2 diabetes mellitus (T2DM). We used restricted cubic spline method and logistic regression method to assess the relationship between CIE L*a*b* color value of tongue and T2DM. A total of 2439 participants (991 T2DM and 1448 healthy) were included. A questionnaire survey and tongue images obtained with tongue diagnosis analysis-1 were analyzed. As required, chi-square and t tests were applied to compare the T2DM and healthy categories. Our findings suggest the 95% confidence interval and odds ratio for body mass index, hypertension, and age were 0.670 (0.531-0.845), 13.461 (10.663-16.993), and 2.595 (2.324-2.897), respectively, when compared to the healthy group. A linear dose-response relationship with an inverse U-shape was determined between CIE L* and CIE a* values and T2DM (P < .001 for overall and P < .001 for nonlinear). Furthermore, U-shaped and linear dose-response associations were identified between T2DM and CIE b* values (P = .0160 for nonlinear). Additionally, in adults, the CIE L*a*b* color value had a correlation with T2DM. This novel perspective provides a multidimensional understanding of traditional Chinese medicine tongue color, elucidating the potential of CIE L*a*b* color values of tongue in the diagnosis of T2DM.

Development and Application of Traditional Chinese Medicine Using AI Machine Learning and Deep Learning Strategies

Traditional Chinese medicine (TCM) has been used for thousands of years and has been proven to be effective at treating many complicated illnesses with minimal side effects. The application and advancement of TCM are, however, constrained by the absence of objective measuring standards due to its relatively abstract diagnostic methods and syndrome differentiation theories. Ongoing developments in machine learning (ML) and deep learning (DL), specifically in computer vision (CV) and natural language processing (NLP), offer novel opportunities to modernize TCM by exploring the profound connotations of its theory. This review begins with an overview of the ML and DL methods employed in TCM; this is followed by practical instances of these applications. Furthermore, extensive discussions emphasize the mature integration of ML and DL in TCM, such as tongue diagnosis, pulse diagnosis, and syndrome differentiation treatment, highlighting their early successful application in the TCM field. Finally, this study validates the accomplishments and addresses the problems and challenges posed by the application and development of TCM powered by ML and DL. As ML and DL techniques continue to evolve, modern technology will spark new advances in TCM.

Multi-feature, Chinese-Western medicine-integrated prediction model for diabetic peripheral neuropathy based on machine learning and SHAP

BACKGROUND

Clinical studies have shown that diabetic peripheral neuropathy (DPN) has been on the rise, with most patients presenting with severe and progressive symptoms. Currently, most of the available prediction models for DPN are derived from general clinical information and laboratory indicators. Several Traditional Chinese medicine (TCM) indicators have been utilised to construct prediction models. In this study, we established a novel machine learning-based multi-featured Chinese-Western medicine-integrated prediction model for DPN using clinical features of TCM.

MATERIALS AND METHODS

The clinical data of 1581 patients with Type 2 diabetes mellitus (T2DM) treated at the Department of Endocrinology of the First Affiliated Hospital of Anhui University of Chinese Medicine were collected. The data (including general information, laboratory parameters and TCM features) of 1142 patients with T2DM were selected after data cleaning. After baseline description analysis of the variables, the data were divided into training and validation sets. Four prediction models were established and their performance was evaluated using validation sets. Meanwhile, the accuracy, precision, recall, F1 score and area under the curve (AUC) of ROC were calculated using ten-fold cross-validation to further assess the performance of the models. An explanatory analysis of the results of the DPN prediction model was carried out using the SHAP framework based on machine learning-based prediction models.

RESULTS

Of the 1142 patients with T2DM, 681 had a comorbidity of DPN, while 461 did not. There was a significant difference between the two groups in terms of age, cause of disease, systolic pressure, HbA1c, ALT, RBC, Cr, BUN, red blood cells in the urine, glucose in the urine, and protein in the urine (p < 0.05). T2DM patients with a comorbidity of DPN exhibited diverse TCM symptoms, including limb numbness, limb pain, hypodynamia, thirst with desire for drinks, dry mouth and throat, blurred vision, gloomy complexion, and unsmooth pulse, with statistically significant differences (p < 0.05). Our results showed that the proposed multi-featured Chinese-Western medicine-integrated prediction model was superior to conventional models without characteristic TCM indicators. The model showed the best performance (accuracy = 0.8109, precision = 0.8029, recall = 0.9060, F1 score = 0.8511, and AUC = 0.9002). SHAP analysis revealed that the dominant risk factors that caused DPN were TCM symptoms (limb numbness, thirst with desire for drinks, blurred vision), age, cause of disease, and glycosylated haemoglobin. These risk factors were exerted positive effects on the DPN prediction models.

CONCLUSIONS

A multi-feature, Chinese-Western medicine-integrated prediction model for DPN was established and validated. The model improves early-stage identification of high-risk groups for DPN in the diagnosis and treatment of T2DM, while also providing informative support for the intelligent management of chronic conditions such as diabetes.

Predictive value of TCM tongue characteristics for chemotherapy-induced myelosuppression in patients with lung cancer

This study aimed to investigate the clinical predictors, including traditional Chinese medicine tongue characteristics and other clinical parameters for chemotherapy-induced myelosuppression (CIM), and then to develop a clinical prediction model and construct a nomogram. A total of 103 patients with lung cancer were prospectively enrolled in this study. All of them were scheduled to receive first-line chemotherapy regimens. Participants were randomly assigned to either the training group (n = 52) or the test group (n = 51). Tongue characteristics and clinical parameters were collected before the start of chemotherapy, and then the incidence of myelosuppression was assessed after treatment. We used univariate logistic regression analysis to identify the risk predictors for assessing the incidence of CIM. Moreover, we developed a predictive model and a nomogram using multivariate logistic regression analysis. Finally, we evaluated the predictive performance of the model by examining the area under the curve value of the receiver operating characteristic, calibration curve, and decision curve analysis. As a result, a total of 3 independent predictors were found to be associated with the CIM in multivariate regression analysis: the fat tongue (OR = 3.67), Karnofsky performance status score (OR = 0.11), and the number of high-toxic drugs in chemotherapy regimens (OR = 4.78). Then a model was constructed using these 3 predictors and it exhibited a robust predictive performance with an area under the curve of 0.82 and the consistent calibration curves. Besides, the decision curve analysis results suggested that applying this predictive model can result in more net clinical benefit for patients. We established a traditional Chinese medicine prediction model based on the tongue characteristics and clinical parameters, which could serve as a useful tool for assessing the risk of CIM.

Artificial intelligence in tongue diagnosis: classification of tongue lesions and normal tongue images using deep convolutional neural network

OBJECTIVE

This study aims to classify tongue lesion types using tongue images utilizing Deep Convolutional Neural Networks (DCNNs).

METHODS

A dataset consisting of five classes, four tongue lesion classes (coated, geographical, fissured tongue, and median rhomboid glossitis), and one healthy/normal tongue class, was constructed using tongue images of 623 patients who were admitted to our clinic. Classification performance was evaluated on VGG19, ResNet50, ResNet101, and GoogLeNet networks using fusion based majority voting (FBMV) approach for the first time in the literature.

RESULTS

In the binary classification problem (normal vs. tongue lesion), the highest classification accuracy performance of 93,53% was achieved utilizing ResNet101, and this rate was increased to 95,15% with the application of the FBMV approach. In the five-class classification problem of tongue lesion types, the VGG19 network yielded the best accuracy rate of 83.93%, and the fusion approach improved this rate to 88.76%.

CONCLUSION

The obtained test results showed that tongue lesions could be identified with a high accuracy by applying DCNNs. Further improvement of these results has the potential for the use of the proposed method in clinic applications.

A review of traditional Chinese medicine diagnosis using machine learning: Inspection, auscultation-olfaction, inquiry, and palpation

Traditional Chinese medicine (TCM) is an essential part of the Chinese medical system and is recognized by the World Health Organization as an important alternative medicine. As an important part of TCM, TCM diagnosis is a method to understand a patient's illness, analyze its state, and identify syndromes. In the long-term clinical diagnosis practice of TCM, four fundamental and effective diagnostic methods of inspection, auscultation-olfaction, inquiry, and palpation (IAOIP) have been formed. However, the diagnostic information in TCM is diverse, and the diagnostic process depends on doctors' experience, which is subject to a high-level subjectivity. At present, the research on the automated diagnosis of TCM based on machine learning is booming. Machine learning, which includes deep learning, is an essential part of artificial intelligence (AI), which provides new ideas for the objective and AI-related research of TCM. This paper aims to review and summarize the current research status of machine learning in TCM diagnosis. First, we review some key factors for the application of machine learning in TCM diagnosis, including data, data preprocessing, machine learning models, and evaluation metrics. Second, we review and summarize the research and applications of machine learning methods in TCM IAOIP and the synthesis of the four diagnostic methods. Finally, we discuss the challenges and research directions of using machine learning methods for TCM diagnosis.

Study on the difference and regularity of tongue images in 309 patients with different pathological stages of non-small cell lung cancer

BACKGROUND

Tongue diagnosis is a crucial traditional Chinese medicine (TCM) inspection method for TCM syndrome differentiation and treatment.

OBJECTIVE

The primary research focus was on tongue image characteristic parameters of patients with non-small cell lung cancer (NSCLC). Analysis of the tongue image parameters of various pathological stages of NSCLC provides technical support for establishing an integrated Chinese and Western auxiliary diagnosis and efficacy evaluation medicine system for lung cancer that integrates tongue image features.

METHODS

Tongue image characteristics of 309 patients with NSCLC and 206 controls were collected and analyzed clinically. The T-test or rank sum test and logistic regression analysis were applied to analyze the characteristics of tongue image indicators of different pathological stages of NSCLC.

RESULTS

There were differences in tongue image characteristics in the NSCLC group compared to the control group. The tongue quality and brightness of the tongue coating in the NSCLC group increased, the red component was reduced, the tongue coating thickened, and the yellow component increased compared to the healthy control group. A comparison of tongue image indexes of NSCLC in different pathological stages showed that stage IV had lower TB-b and higher TB-a than stage I. In addition, stage IV had lower TB-b than stage II + III, showing an increase in the blue and red components of the tongue in stage IV and the appearance of cyanotic tongue features.

CONCLUSION

The tongue image characteristics of NSCLC patients differed from those of the control group. Tongue imaging indicators can reflect the characteristics of tongue images of patients with NSCLC. The tongue image characteristics of patients with stage IV lung cancer are bluish and purple compared with those with stage I, II, and III. It is suggested that the tongue's image characteristics can be used as a reference for the pathological classification of NSCLC and judgment of the disease process.

Core characteristics of sublingual veins analysis and its relationship with hypertension

BACKGROUND

The sublingual vein (SV) is a specialized diagnostic method used in Traditional Chinese Medicine (TCM). Despite its ability to objectively reflect blood flow, SV is often overlooked in clinical practice.

OBJECTIVE

This study aims to analyze the core characteristics of SV and investigate the in-depth relationship between its digital characteristics and hypertension. The goal is to find a link between SV and hypertension and break out of the current situation.

METHODS

Modern digital analysis techniques were applied to the traditional SV diagnostic theory. In a controlled study with 204 participants, the digital characteristics of SV were documented using TFDA-1, and its color value was analyzed using TDAS. Morphological characteristics of SV, such as trunklength, width, and tortuosity, were examined by combining computer vision with expert interpretation. This involved the application of automatic ranging methods and a rectangular approximation algorithm, which are novel approaches in the field of TCM. The t-test and Mann-Whitney U test were used to analyze the digital characteristics of SV in hypertension. Binary logistic regression and neural network models were established using machine learning to explore the deep relationship between SV characteristics and hypertension.

RESULTS

There was a significant difference of the tortuosity of SV between the two groups (Z=-2.629, p= 0.009). The results revealed thick width of SV (OR = 2.64, 95% CI: 1.02-6.79) was the risk factor for hypertension. Addition of SV characteristics improved overall percent correct for hypertension prediction to 80%.

CONCLUSION

TCM method of diagnosis of SV has been greatly expanded in terms of technical means, and the close relationship between SV and hypertension has been found in clinical data.

ODQN-Net: Optimized Deep Q Neural Networks for Disease Prediction Through Tongue Image Analysis Using Remora Optimization Algorithm

Tongue analysis plays the major role in disease type prediction and classification according to Indian ayurvedic medicine. Traditionally, there is a manual inspection of tongue image by the expert ayurvedic doctor to identify or predict the disease. However, this is time-consuming and even imprecise. Due to the advancements in recent machine learning models, several researchers addressed the disease prediction from tongue image analysis. However, they have failed to provide enough accuracy. In addition, multiclass disease classification with enhanced accuracy is still a challenging task. Therefore, this article focuses on the development of optimized deep q-neural network (DQNN) for disease identification and classification from tongue images, hereafter referred as ODQN-Net. Initially, the multiscale retinex approach is introduced for enhancing the quality of tongue images, which also acts as a noise removal technique. In addition, a local ternary pattern is used to extract the disease-specific and disease-dependent features based on color analysis. Then, the best features are extracted from the available features set using the natural inspired Remora optimization algorithm with reduced computational time. Finally, the DQNN model is used to classify the type of diseases from these pretrained features. The obtained simulation performance on tongue imaging data set proved that the proposed ODQN-Net resulted in superior performance compared with state-of-the-art approaches with 99.17% of accuracy and 99.75% and 99.84% of F1-score and Mathew's correlation coefficient, respectively.

Machine learning prediction models for different stages of non-small cell lung cancer based on tongue and tumor marker: a pilot study

OBJECTIVE

To analyze the tongue feature of NSCLC at different stages, as well as the correlation between tongue feature and tumor marker, and investigate the feasibility of establishing prediction models for NSCLC at different stages based on tongue feature and tumor marker.

METHODS

Tongue images were collected from non-advanced NSCLC patients (n = 109) and advanced NSCLC patients (n = 110), analyzed the tongue images to obtain tongue feature, and analyzed the correlation between tongue feature and tumor marker in different stages of NSCLC. On this basis, six classifiers, decision tree, logistic regression, SVM, random forest, naive bayes, and neural network, were used to establish prediction models for different stages of NSCLC based on tongue feature and tumor marker.

RESULTS

There were statistically significant differences in tongue feature between the non-advanced and advanced NSCLC groups. In the advanced NSCLC group, the number of indexes with statistically significant correlations between tongue feature and tumor marker was significantly higher than in the non-advanced NSCLC group, and the correlations were stronger. Support Vector Machine (SVM), decision tree, and logistic regression among the machine learning methods performed poorly in models with different stages of NSCLC. Neural network, random forest and naive bayes had better classification efficiency for the data set of tongue feature and tumor marker and baseline. The models' classification accuracies were 0.767 ± 0.081, 0.718 ± 0.062, and 0.688 ± 0.070, respectively, and the AUCs were 0.793 ± 0.086, 0.779 ± 0.075, and 0.771 ± 0.072, respectively.

CONCLUSIONS

There were statistically significant differences in tongue feature between different stages of NSCLC, with advanced NSCLC tongue feature being more closely correlated with tumor marker. Due to the limited information, single data sources including baseline, tongue feature, and tumor marker cannot be used to identify the different stages of NSCLC in this pilot study. In addition to the logistic regression method, other machine learning methods, based on tumor marker and baseline data sets, can effectively improve the differential diagnosis efficiency of different stages of NSCLC by adding tongue image data, which requires further verification based on large sample studies in the future.

Relationship chains of subhealth physical examination indicators: a cross-sectional study using the PLS-SEM approach

Subhealth is a transitional state between health and disease, and it can be detected through routine physical check-ups. However, the complexity and diversity of physical examination items and the difficulty of quantifying subhealth manifestations are the main problems that hinder its treatment. The aim of this study was to systematically investigate the physical examination performance of the subhealthy population and further explore the deeper relationships between indicators. Indicators were obtained for 878 subjects, including basic information, Western medicine indicators, inquiries of traditional Chinese medicine and sublingual vein (SV) characteristics. Statistical differences were analysed using R software. To explore the distribution of symptoms and symptom clusters in subhealth, partial least squares-structural equation modelling (PLS-SEM) was applied to the subhealth physical examination index, and a structural model was developed to verify whether the relationship chain between the latent variables was reasonable. Finally, the reliability and validity of the PLS-SE model were assessed. The most common subclinical clinical symptoms were limb soreness (37.6%), fatigue (31.6%), shoulder and neck pain (30.5%) and dry eyes (29.2%). The redness of the SV in the subhealthy group was paler than that in the healthy group (p < 0.001). This study validates the establishment of the directed acyclic relationship chain in the subhealthy group: the path from routine blood tests to lipid metabolism (t = 7.878, p < 0.001), the path from lipid metabolism to obesity (t = 8.410, p < 0.001), the path from obesity to SV characteristics (t = 2.237, p = 0.025), and the path from liver function to SV characteristics (t = 2.215, p = 0.027). The innovative application of PLS-SEM to the study of subhealth has revealed the existence of a chain of relationships between physical examination indicators, which will provide a basis for further exploration of subhealth mechanisms and causal inference. This study has identified the typical symptoms of subhealth, and their early management will help to advance the treatment of diseases.

A lung cancer risk warning model based on tongue images

Objective: To investigate the tongue image features of patients with lung cancer and benign pulmonary nodules and to construct a lung cancer risk warning model using machine learning methods. Methods: From July 2020 to March 2022, we collected 862 participants including 263 patients with lung cancer, 292 patients with benign pulmonary nodules, and 307 healthy subjects. The TFDA-1 digital tongue diagnosis instrument was used to capture tongue images, using feature extraction technology to obtain the index of the tongue images. The statistical characteristics and correlations of the tongue index were analyzed, and six machine learning algorithms were used to build prediction models of lung cancer based on different data sets. Results: Patients with benign pulmonary nodules had different statistical characteristics and correlations of tongue image data than patients with lung cancer. Among the models based on tongue image data, the random forest prediction model performed the best, with a model accuracy of 0.679 ± 0.048 and an AUC of 0.752 ± 0.051. The accuracy for the logistic regression, decision tree, SVM, random forest, neural network, and naïve bayes models based on both the baseline and tongue image data were 0.760 ± 0.021, 0.764 ± 0.043, 0.774 ± 0.029, 0.770 ± 0.050, 0.762 ± 0.059, and 0.709 ± 0.052, respectively, while the corresponding AUCs were 0.808 ± 0.031, 0.764 ± 0.033, 0.755 ± 0.027, 0.804 ± 0.029, 0.777 ± 0.044, and 0.795 ± 0.039, respectively. Conclusion: The tongue diagnosis data under the guidance of traditional Chinese medicine diagnostic theory was useful. The performance of models built on tongue image and baseline data was superior to that of the models built using only the tongue image data or the baseline data. Adding objective tongue image data to baseline data can significantly improve the efficacy of lung cancer prediction models.

Predictive value of TCM clinical index for diabetic peripheral neuropathy among the type 2 diabetes mellitus population: A new observation and insight

Aims

The objectives of this study were to identify clinical predictors of the Traditional Chinese medicine (TCM) clinical index for diabetic peripheral neuropathy (DPN) in type 2 diabetes mellitus (T2DM) patients, develop a clinical prediction model, and construct a nomogram.

Methods

We collected the TCM clinical index from 3590 T2DM recruited at the Second Affiliated Hospital of Tianjin University of Traditional Chinese Medicine from January 2019 to October 2020. The participants were randomly assigned to either the training group (n = 3297) or the validation group (n = 1426). TCM symptoms and tongue characteristics were used to assess the risk of developing DPN in T2DM patients. Through 5-fold cross-validation in the training group, the least absolute shrinkage and selection operator (LASSO) regression analysis method was used to optimize variable selection. In addition, using multifactor logistic regression analysis, a predictive model and nomogram were developed.

Results

A total of eight independent predictors were found to be associated with the DPN in multivariate logistic regression analyses: advanced age of grading (odds ratio/OR 1.575), smoke (OR 2.815), insomnia (OR 0.557), sweating (OR 0.535), loose teeth (OR 1.713), dry skin (OR 1.831), purple tongue (OR 2.278). And dark red tongue (OR 0.139). The model was constructed using these eight predictor's medium discriminative capabilities. The area under the curve (AUC) of the training set is 0.727, and the AUC of the validation set is 0.744 on the ROC curve. The calibration plot revealed that the model's goodness-of-fit is satisfactory.

Conclusions

We established a TCM prediction model for DPN in patients with T2DM based on the TCM clinical index.

Performance of a prediabetes risk prediction model: A systematic review

Backgrounds

The prediabetes population is large and easily overlooked because of the lack of obvious symptoms, which can progress to diabetes. Early screening and targeted interventions can substantially reduce the rate of conversion of prediabetes to diabetes. Therefore, this study systematically reviewed prediabetes risk prediction models, performed a summary and quality evaluation, and aimed to recommend the optimal model.

Methods

We systematically searched five databases (Cochrane, PubMed, Embase, Web Of Science, and CNKI) for published literature related to prediabetes risk prediction models and excluded preprints, duplicate publications, reviews, editorials, and other studies, with a search time frame of March 01, 2023. Data were categorized and summarized using a standardized data extraction form that extracted data including author; publication date; study design; country; demographic characteristics; assessment tool name; sample size; study type; and model-related indicators. The PROBAST tool was used to assess the risk of bias profile of included studies.

Findings

14 studies with a total of 15 models were eventually included in the systematic review. We found that the most common predictors of models were age, family history of diabetes, gender, history of hypertension, and BMI. Most of the studies (83.3%) had a high risk of bias, mainly related to under-reporting of outcome information and poor methodological design during the development and validation of models. Due to the low quality of included studies, the evidence for predictive validity of the available models is unclear.

Interpretation

We should pay attention to the early screening of prediabetes patients and give timely pharmacological and lifestyle interventions. The predictive performance of the existing model is not satisfactory, and the model building process can be standardized and external validation can be added to improve the accuracy of the model in the future.

KFPredict: An ensemble learning prediction framework for diabetes based on fusion of key features

BACKGROUND AND OBJECTIVE

Diabetes is a disease that requires early detection and early treatment, and complications are likely to occur in late stages of the disease, threatening the life of patients. Therefore, in order to diagnose diabetic patients as early as possible, it is necessary to establish a model that can accurately predict diabetes.

METHODOLOGY

This paper proposes an ensemble learning framework: KFPredict, which combines multi-input models with key features and machine learning algorithms. We first propose a multi-input neural network model (KF_NN) that fuses key features and uses a decision tree-based selection recursive feature elimination algorithm and correlation coefficient method to screen out the key feature inputs and secondary feature inputs in the model. We then ensemble KF_NN with three machine learning algorithms (i.e., Support Vector Machine, Random Forest and K-Nearest Neighbors) for soft voting to form our predictive classifier for diabetes prediction.

RESULTS

Our framework demonstrates good prediction results on the test set with a sensitivity of 0.85, a specificity of 0.98, and an accuracy of 93.5%. Compared with the single prediction method KFPredict, the accuracy is up to 18.18% higher. Concurrently, we also compared KFPredict with the existing prediction methods. It still has good prediction performance, and the accuracy rate is improved by up to 14.93%.

CONCLUSION

This paper constructs a diabetes prediction framework that combines multi-input models with key features and machine learning algorithms. Taking tthe PIMA diabetes dataset as the test data, the experiment shows that the framework presents good prediction results.

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.

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.

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.

A novel multistage ensemble approach for prediction and classification of diabetes

Diabetes mellitus is a metabolic syndrome affecting millions of people worldwide. Every year, the rate of occurrence rises drastically. Diabetes-related problems across several vital organs of the body can be fatal if left untreated. Diabetes must be detected early to receive proper treatment, preventing the condition from escalating to severe problems. Tremendous health sciences and biotechnology advancements have resulted in massive data that generated massive Electronic Health Records and clinical information. The exponential increase of electronically gathered information has resulted in more complicated, accurate prediction models that can be updated continuously using machine learning techniques. This research mainly emphasizes discovering the best ensemble model for predicting diabetes. A new multistage ensemble model is proposed for diabetes prediction. In this model, accuracy is predicated on the Pima Indian Diabetes dataset. The accuracy of the proposed ensemble model is compared with the existing machine learning model, and the experimental results demonstrate the performance of the proposed model in terms of higher Precision, f-measure, Recall, and area under the curve.

Chinese Medicine Phenomics (Chinmedphenomics): Personalized, Precise and Promising

The systematicness of phenomics and Traditional Chinese Medicine (TCM) enable these two disciplines to interlink with each other. This article discussed the similarity in theory and application between TCM and phenomics and illustrates their respective advantages in diagnosis and treatment of diseases, forming a new discipline eventually. Chinese medicine phenomics (Chinmedphenomics) is built on classic TCM, combined with phenomics technology, and the development of which needs the mega cohort with TCM syndrome and the characteristics of precision medicine as well as multi-disciplinary cooperation, which is personalized, precise and promising, providing unique scientific insights into understanding human health.

Deep Learning Multi-label Tongue Image Analysis and Its Application in a Population Undergoing Routine Medical Checkup

Background

Research on intelligent tongue diagnosis is a main direction in the modernization of tongue diagnosis technology. Identification of tongue shape and texture features is a difficult task for tongue diagnosis in traditional Chinese medicine (TCM). This study aimed to explore the application of deep learning techniques in tongue image analyses.

Methods

A total of 8676 tongue images were annotated by clinical experts, into seven categories, including the fissured tongue, tooth-marked tongue, stasis tongue, spotted tongue, greasy coating, peeled coating, and rotten coating. Based on the labeled tongue images, the deep learning model faster region-based convolutional neural networks (Faster R-CNN) was utilized to classify tongue images. Four performance indices, i.e., accuracy, recall, precision, and F1-score, were selected to evaluate the model. Also, we applied it to analyze tongue image features of 3601 medical checkup participants in order to explore gender and age factors and the correlations among tongue features in diseases through complex networks.

Results

The average accuracy, recall, precision, and F1-score of our model achieved 90.67%, 91.25%, 99.28%, and 95.00%, respectively. Over the tongue images from the medical checkup population, the model Faster R-CNN detected 41.49% fissured tongue images, 37.16% tooth-marked tongue images, 29.66% greasy coating images, 18.66% spotted tongue images, 9.97% stasis tongue images, 3.97% peeled coating images, and 1.22% rotten coating images. There were significant differences in the incidence of the fissured tongue, tooth-marked tongue, spotted tongue, and greasy coating among age and gender. Complex networks revealed that fissured tongue and tooth-marked were closely related to hypertension, dyslipidemia, overweight and nonalcoholic fatty liver disease (NAFLD), and a greasy coating tongue was associated with hypertension and overweight.

Conclusion

The model Faster R-CNN shows good performance in the tongue image classification. And we have preliminarily revealed the relationship between tongue features and gender, age, and metabolic diseases in a medical checkup population.

Automatic tongue image quality assessment using a multi-task deep learning model

The quality of tongue images has a significant influence on the performance of tongue diagnosis in Chinese medicine. During the acquisition process, the quality of the tongue image is easily affected by factors such as the illumination, camera parameters, and tongue extension of the subject. To ensure that the quality of the collected images meet the diagnostic criteria of traditional Chinese Medicine practitioners, we propose a deep learning model to evaluate the quality of tongue images. First, we acquired the tongue images of the patients under different lighting conditions, exposures, and tongue extension conditions using the inspection instrument, and experienced Chinese physicians manually screened them into high-quality and unqualified tongue datasets. We then designed a multi-task deep learning network to classify and evaluate the quality of tongue images by adding tongue segmentation as an auxiliary task, as the two tasks are related and can promote each other. Finally, we adaptively designed different task weight coefficients of a multi-task network to obtain better tongue image quality assessment (IQA) performance, as the two tasks have relatively different contributions in the loss weighting scheme. Experimental results show that the proposed method is superior to the traditional deep learning tongue IQA method, and as an additional task of the network, it can output the tongue segmentation area, which provides convenience for follow-up clinical tongue diagnosis. In addition, we used network visualization to verify the effectiveness of the proposed method qualitatively.

Hybrid stacked ensemble combined with genetic algorithms for diabetes prediction

Diabetes is currently one of the most common, dangerous, and costly diseases globally caused by increased blood sugar or a decrease in insulin in the body. Diabetes can have detrimental effects on people’s health if diagnosed late. Today, diabetes has become one of the challenges for health and government officials. Prevention is a priority, and taking care of people’s health without compromising their comfort is an essential need. In this study, the ensemble training methodology based on genetic algorithms was used to diagnose and predict the outcomes of diabetes mellitus accurately. This study uses the experimental data, actual data on Indian diabetics on the University of California website. Current developments in ICT, such as the Internet of Things, machine learning, and data mining, allow us to provide health strategies with more intelligent capabilities to accurately predict the outcomes of the disease in daily life and the hospital and prevent the progression of this disease and its many complications. The results show the high performance of the proposed method in diagnosing the disease, which has reached 98.8%, and 99% accuracy in this study.

A multi-step approach for tongue image classification in patients with diabetes

BACKGROUND

In China, diabetes is a common, high-incidence chronic disease. Diabetes has become a severe public health problem. However, the current diagnosis and treatment methods are difficult to control the progress of diabetes. Traditional Chinese Medicine (TCM) has become an option for the treatment of diabetes due to its low cost, good curative effect, and good accessibility.

OBJECTIVE

Based on the tongue images data to realize the fine classification of the diabetic population, provide a diagnostic basis for the formulation of individualized treatment plans for diabetes, ensure the accuracy and consistency of the TCM diagnosis, and promote the objective and standardized development of TCM diagnosis.

METHODS

We use the TFDA-1 tongue examination instrument to collect the tongue images of the subjects. Tongue Diagnosis Analysis System (TDAS) is used to extract the TDAS features of the tongue images. Vector Quantized Variational Autoencoder (VQ-VAE) extracts VQ-VAE features from tongue images. Based on VQ-VAE features, K-means clustering tongue images. TDAS features are used to describe the differences between clusters. Vision Transformer (ViT) combined with Grad-weighted Class Activation Mapping (Grad-CAM) is used to verify the clustering results and calculate positioning diagnostic information.

RESULTS

Based on VQ-VAE features, K-means divides the diabetic population into 4 clusters with clear boundaries. The silhouette, calinski harabasz, and davies bouldin scores are 0.391, 673.256, and 0.809, respectively. Cluster 1 had the highest Tongue Body L (TB-L) and Tongue Coating L (TC-L) and the lowest Tongue Coating Angular second moment (TC-ASM), with a pale red tongue and white coating. Cluster 2 had the highest TC-b with a yellow tongue coating. Cluster 3 had the highest TB-a with a red tongue. Group 4 had the lowest TB-L, TC-L, and TB-b and the highest Per-all with a purple tongue and the largest tongue coating area. ViT verifies the clustering results of K-means, the highest Top-1 Classification Accuracy (CA) is 87.8%, and the average CA is 84.4%.

CONCLUSIONS

The study organically combined unsupervised learning, self-supervised learning, and supervised learning and designed a complete diabetic tongue image classification method. This method does not rely on human intervention, makes decisions based entirely on tongue image data, and achieves state-of-the-art results. Our research will help TCM deeply participate in the individualized treatment of diabetes and provide new ideas for promoting the standardization of TCM diagnosis.

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.

A novel early diagnostic framework for chronic diseases with class imbalance

Chronic diseases are one of the most severe health issues in the world, due to their terrible clinical presentations such as long onset cycle, insidious symptoms, and various complications. Recently, machine learning has become a promising technique to assist the early diagnosis of chronic diseases. However, existing works ignore the problems of feature hiding and imbalanced class distribution in chronic disease datasets. In this paper, we present a universal and efficient diagnostic framework to alleviate the above two problems for diagnosing chronic diseases timely and accurately. Specifically, we first propose a network-limited polynomial neural network (NLPNN) algorithm to efficiently capture high-level features hidden in chronic disease datasets, which is data augmentation in terms of its feature space and can also avoid over-fitting. Then, to alleviate the class imbalance problem, we further propose an attention-empowered NLPNN algorithm to improve the diagnostic accuracy for sick cases, which is also data augmentation in terms of its sample space. We evaluate the proposed framework on nine public and two real chronic disease datasets (partly with class imbalance). Extensive experiment results demonstrate that the proposed diagnostic algorithms outperform state-of-the-art machine learning algorithms, and can achieve superior performances in terms of accuracy, recall, F1, and G_mean. The proposed framework can help to diagnose chronic diseases timely and accurately at an early stage.

A Study of Logistic Regression for Fatigue Classification Based on Data of Tongue and Pulse

Methods

The Tongue and Face Diagnosis Analysis-1 instrument and Pulse Diagnosis Analysis-1 instrument were used to collect the tongue image and sphygmogram of the subhealth fatigue population (n = 252) and disease fatigue population (n = 1160), and we mainly analyzed the tongue and pulse characteristics and constructed the classification model by using the logistic regression method.

Results

The results showed that subhealth fatigue people and disease fatigue people had different characteristics of tongue and pulse, and the logistic regression model based on tongue and pulse data had a good classification effect. The accuracies of models of healthy controls and subhealth fatigue, subhealth fatigue and disease fatigue, and healthy controls and disease fatigue were 68.29%, 81.18%, and 84.73%, and the AUC was 0.698, 0.882, and 0.924, respectively.

Conclusion

This study provided a new noninvasive method for the fatigue diagnosis from the perspective of objective tongue and pulse data, and the modern tongue diagnosis and pulse diagnosis have good application prospects.

A New Approach of Fatigue Classification Based on Data of Tongue and Pulse With Machine Learning

BACKGROUND

Fatigue is a common and subjective symptom, which is associated with many diseases and suboptimal health status. A reliable and evidence-based approach is lacking to distinguish disease fatigue and non-disease fatigue. This study aimed to establish a method for early differential diagnosis of fatigue, which can be used to distinguish disease fatigue from non-disease fatigue, and to investigate the feasibility of characterizing fatigue states in a view of tongue and pulse data analysis.

METHODS

Tongue and Face Diagnosis Analysis-1 (TFDA-1) instrument and Pulse Diagnosis Analysis-1 (PDA-1) instrument were used to collect tongue and pulse data. Four machine learning models were used to perform classification experiments of disease fatigue vs. non-disease fatigue.

RESULTS

The results showed that all the four classifiers over "Tongue & Pulse" joint data showed better performances than those only over tongue data or only over pulse data. The model accuracy rates based on logistic regression, support vector machine, random forest, and neural network were (85.51 ± 1.87)%, (83.78 ± 4.39)%, (83.27 ± 3.48)% and (85.82 ± 3.01)%, and with Area Under Curve estimates of 0.9160 ± 0.0136, 0.9106 ± 0.0365, 0.8959 ± 0.0254 and 0.9239 ± 0.0174, respectively.

CONCLUSION

This study proposed and validated an innovative, non-invasive differential diagnosis approach. Results suggest that it is feasible to characterize disease fatigue and non-disease fatigue by using objective tongue data and pulse data.

Panoramic tongue imaging and deep convolutional machine learning model for diabetes diagnosis in humans

Diabetes is a serious metabolic disorder with high rate of prevalence worldwide; the disease has the characteristics of improper secretion of insulin in pancreas that results in high glucose level in blood. The disease is also associated with other complications such as cardiovascular disease, retinopathy, neuropathy and nephropathy. The development of computer aided decision support system is inevitable field of research for disease diagnosis that will assist clinicians for the early prognosis of diabetes and to facilitate necessary treatment at the earliest. In this research study, a Traditional Chinese Medicine based diabetes diagnosis is presented based on analyzing the extracted features of panoramic tongue images such as color, texture, shape, tooth markings and fur. The feature extraction is done by Convolutional Neural Network (CNN)—ResNet 50 architecture, and the classification is performed by the proposed Deep Radial Basis Function Neural Network (RBFNN) algorithm based on auto encoder learning mechanism. The proposed model is simulated in MATLAB environment and evaluated with performance metrics—accuracy, precision, sensitivity, specificity, F1 score, error rate, and receiver operating characteristics (ROC). On comparing with existing models, the proposed CNN based Deep RBFNN machine learning classifier model outperformed with better classification performance and proving its effectiveness.

Development and validation of a machine learning-augmented algorithm for diabetes screening in community and primary care settings: A population-based study

BACKGROUND

Opportunely screening for diabetes is crucial to reduce its related morbidity, mortality, and socioeconomic burden. Machine learning (ML) has excellent capability to maximize predictive accuracy. We aim to develop ML-augmented models for diabetes screening in community and primary care settings.

METHODS

8425 participants were involved from a population-based study in Hubei, China since 2011. The dataset was split into a development set and a testing set. Seven different ML algorithms were compared to generate predictive models. Non-laboratory features were employed in the ML model for community settings, and laboratory test features were further introduced in the ML+lab models for primary care. The area under the receiver operating characteristic curve (AUC), area under the precision-recall curve (auPR), and the average detection costs per participant of these models were compared with their counterparts based on the New China Diabetes Risk Score (NCDRS) currently recommended for diabetes screening.

RESULTS

The AUC and auPR of the ML model were 0·697and 0·303 in the testing set, seemingly outperforming those of NCDRS by 10·99% and 64·67%, respectively. The average detection cost of the ML model was 12·81% lower than that of NCDRS with the same sensitivity (0·72). Moreover, the average detection cost of the ML+FPG model is the lowest among the ML+lab models and less than that of the ML model and NCDRS+FPG model.

CONCLUSION

The ML model and the ML+FPG model achieved higher predictive accuracy and lower detection costs than their counterpart based on NCDRS. Thus, the ML-augmented algorithm is potential to be employed for diabetes screening in community and primary care settings.

Artificial intelligence and diabetes technology: A review

Artificial intelligence (AI) is widely discussed in the popular literature and is portrayed as impacting many aspects of human life, both in and out of the workplace. The potential for revolutionizing healthcare is significant because of the availability of increasingly powerful computational platforms and methods, along with increasingly informative sources of patient data, both in and out of clinical settings. This review aims to provide a realistic assessment of the potential for AI in understanding and managing diabetes, accounting for the state of the art in the methodology and medical devices that collect data, process data, and act accordingly. Acknowledging that many conflicting definitions of AI have been put forth, this article attempts to characterize the main elements of the field as they relate to diabetes, identifying the main perspectives and methods that can (i) affect basic understanding of the disease, (ii) affect understanding of risk factors (genetic, clinical, and behavioral) of diabetes development, (iii) improve diagnosis, (iv) improve understanding of the arc of disease (progression and personal/societal impact), and finally (v) improve treatment.

A New Method for Syndrome Classification of Non-Small-Cell Lung Cancer Based on Data of Tongue and Pulse with Machine Learning

Objective

To explore the data characteristics of tongue and pulse of non-small-cell lung cancer with Qi deficiency syndrome and Yin deficiency syndrome, establish syndrome classification model based on data of tongue and pulse by using machine learning methods, and evaluate the feasibility of syndrome classification based on data of tongue and pulse.

Methods

We collected tongue and pulse of non-small-cell lung cancer patients with Qi deficiency syndrome (n = 163), patients with Yin deficiency syndrome (n = 174), and healthy controls (n = 185) using intelligent tongue diagnosis analysis instrument and pulse diagnosis analysis instrument, respectively. We described the characteristics and examined the correlation of data of tongue and pulse. Four machine learning methods, namely, random forest, logistic regression, support vector machine, and neural network, were used to establish the classification models based on symptom, tongue and pulse, and symptom and tongue and pulse, respectively.

Results

Significant difference indices of tongue diagnosis between Qi deficiency syndrome and Yin deficiency syndrome were TB-a, TB-S, TB-Cr, TC-a, TC-S, TC-Cr, perAll, and the tongue coating texture indices including TC-CON, TC-ASM, TC-MEAN, and TC-ENT. Significant difference indices of pulse diagnosis were t₄ and t₅. The classification performance of each model based on different datasets was as follows: tongue and pulse < symptom < symptom and tongue and pulse. The neural network model had a better classification performance for symptom and tongue and pulse datasets, with an area under the ROC curves and accuracy rate which were 0.9401 and 0.8806.

Conclusions

It was feasible to use tongue data and pulse data as one of the objective diagnostic basis in Qi deficiency syndrome and Yin deficiency syndrome of non-small-cell lung cancer.

Application of computer tongue image analysis technology in the diagnosis of NAFLD

Nonalcoholic fatty liver disease (NAFLD), a leading cause of chronic hepatic disease, can progress to liver fibrosis, cirrhosis, and hepatocellular carcinoma. Therefore, it is extremely important to explore early diagnosis and screening methods. In this study, we developed models based on computer tongue image analysis technology to observe the tongue characteristics of 1778 participants (831 cases of NAFLD and 947 cases of non-NAFLD). Combining quantitative tongue image features, basic information, and serological indexes, including the hepatic steatosis index (HSI) and fatty liver index (FLI), we utilized machine learning methods, including Logistic Regression, Support Vector Machine (SVM), Random Forest (RF), Gradient Boosting Decision Tree (GBDT), Adaptive Boosting Algorithm (AdaBoost), Naïve Bayes, and Neural Network for NAFLD diagnosis. The best fusion model for diagnosing NAFLD by Logistic Regression, which contained the tongue image parameters, waist circumference, BMI, GGT, TG, and ALT/AST, achieved an AUC of 0.897 (95% CI, 0.882-0.911), an accuracy of 81.70% with a sensitivity of 77.62% and a specificity of 85.22%; in addition, the positive likelihood ratio and negative likelihood ratio were 5.25 and 0.26, respectively. The application of computer intelligent tongue diagnosis technology can improve the accuracy of NAFLD diagnosis and may provide a convenient technical reference for the establishment of early screening methods for NAFLD, which is worth further research and verification.