Feasibility of tongue image detection for coronary artery disease: based on deep learning

Frontiers in Cardiovascular Medicine, Journal Year: 2024, Volume and Issue: 11

Published: Aug. 23, 2024

Aim Clarify the potential diagnostic value of tongue images for coronary artery disease (CAD), develop a CAD model that enhances performance by incorporating image inputs, and provide more reliable evidence clinical diagnosis CAD, offering new biological characterization evidence. Methods We recruited 684 patients from four hospitals in China cross-sectional study, collecting their baseline information standardized to train validate our algorithm. used DeepLabV3 + segmentation body employed Resnet-18, pretrained on ImageNet, extract features images. applied DT (Decision Trees), RF (Random Forest), LR (Logistic Regression), SVM (Support Vector Machine), XGBoost models, developing models with inputs risk factors alone then additional inclusion features. compared different algorithms using accuracy, precision, recall, F1-score, AUPR, AUC. Results classified found this classification criterion was effective (ACC = 0.670, AUC 0.690, Recall 0.666). After comparing such as Decision Tree (DT), Random Forest (RF), Logistic Regression (LR), Support Machine (SVM), XGBoost, we ultimately chose The algorithm developed solely based ACC 0.730, Precision 0.811, 0.763. When were integrated, improved 0.760, 0.773, 0.786, 0.850, indicating an enhancement performance. Conclusion use is feasible, these can enhance existing algorithms. have customized novel algorithm, which offers advantages being noninvasive, simple, cost-effective. It suitable large-scale screening among hypertensive populations. Tongue may emerge biomarkers indicators CAD.

Language: Английский

---

Clinical Study of Intelligent Tongue Diagnosis and Oral Microbiome for Classifying TCM Syndromes in MASLD

Published: April 23, 2025

Background This study aimed to analyze the tongue image features and oral microbial markers in different TCM syndromes related metabolic dysfunction-associated steatotic liver disease (MASLD). Methods involved 34 healthy volunteers 66 MASLD patients [36 with Dampness-Heat (DH) 30 Qi-Deficiency (QD) syndrome]. Oral microbiome analysis was conducted through 16S rRNA sequencing. Tongue feature extraction used Uncertainty Augmented Context Attention Network (UACANet), while syndrome classification performed using five machine learning methods based on microbiota. Results Significant differences color, coating, microbiota were noted between DH band QD patients. exhibited a red-crimson color greasy coating enriched Streptococcus Rothia tongue. In contrast, displayed pale higher abundances of Neisseria, Fusobacterium, Porphyromonas Haemophilus. Combining characteristics differentiated an AUC 0.939 accuracy 85%. Conclusion suggests that are metabolism, possess distinct biomarkers, supporting classification.

Language: Английский

---

Research status and prospect of tongue image diagnosis analysis based on machine learning

Digital Chinese Medicine, Journal Year: 2024, Volume and Issue: 7(1), P. 3 - 12

Published: March 1, 2024

Image-based intelligent diagnosis represents a trending research direction in the field of tongue traditional Chinese medicine (TCM). In recent years, machine learning techniques, including convolutional neural networks (CNNs) and Transformers, have been widely used analysis medical images, such as computed tomography (CT) nuclear magnetic resonance imaging (MRI). These techniques significantly enhanced efficiency accuracy decision-making TCM practices. Advanced artificial intelligence (AI) technologies also provided new opportunities for development equipment diagnosis, resulting improved standardization diagnostic procedures. Although image methods could transform images into scientific analyzable data, recognizing analyzing that capture complicated features tooth-marked tongue, spots prickles, fissured variations coating thickness, texture (curdy greasy), presence (peeled coating) continues posing significant challenges contemporary research. Therefore, employment shape well their applications is focus this study. study, both deep were summarized analyzed to figure out value predicting disease risks by observing shapes textures, aiming open chapter application AI short, combination technologies, will not only enhance basis but improve its clinical applicability, thereby advancing modernization therapeutic

Language: Английский

---

Application of tongue image characteristics and oral-gut microbiota in predicting pre-diabetes and type 2 diabetes with machine learning

Frontiers in Cellular and Infection Microbiology, Journal Year: 2024, Volume and Issue: 14

Published: Nov. 4, 2024

Background This study aimed to characterize the oral and gut microbiota in prediabetes mellitus (Pre-DM) type 2 diabetes (T2DM) patients while exploring association between tongue manifestations oral-gut axis progression. Methods Participants included 30 Pre-DM patients, 37 individuals with T2DM, 28 healthy controls. Tongue images oral/fecal samples were analyzed using image processing 16S rRNA sequencing. Machine learning techniques, including support vector machine (SVM), random forest, gradient boosting, adaptive K-nearest neighbors, applied integrate data profiles construct predictive models for T2DM classification. Results Significant shifts characteristics identified during progression from T2DM. Elevated Firmicutes levels along associated white greasy fur, indicative of underlying metabolic changes. An SVM-based model demonstrated an accuracy 78.9%, AUC 86.9%. Notably, parameters (TB-a, perALL) specific ( Escherichia , Porphyromonas-A ) emerged as prominent diagnostic markers Conclusion The integration diagnosis microbiome analysis reveals distinct features microbial markers. approach significantly improves capability

Language: Английский

---

Generative AI: A transformative force in advancing research and care in metabolic dysfunction-associated fatty liver disease

Liver Research, Journal Year: 2024, Volume and Issue: 8(2), P. 127 - 129

Published: May 23, 2024

Language: Английский

---

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

Medicine, Journal Year: 2024, Volume and Issue: 103(25), P. e38575 - e38575

Published: June 21, 2024

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

Language: Английский

---