Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
April 14th, 2023
The present study employed U-Net and other deep learning algorithms to segment a tongue image and compared the segmentation results to investigate the objectification of tongue diagnosis.
This protocol uses deep learning algorithm to segment tongue images and automatically separates the tongue from the whole image, which can be used for further research. This protocol compares the performance of four-tongue image segmentation algorithm to select the base algorithm that can segment a tongue with a clear counter. Patented instruments and algorithm are used in this study.
It is recommended to use professional instruments and master computer technology. To begin, prepare the self-developed handheld lingual face diagnostic instrument to collect lingual face images of patients. Fill in the patient's name, gender, age, and disease on the computer page.
Next, ask the patient to sit upright and position the image acquisition device connected to the computer. Then, position the face of the patient in the image acquisition instrument and ask them to extend their tongue out of their mouth to the maximum extent. Verify through the images on the computer screen that the patient is in the correct position and that the tongue and face are fully exposed.
Press the shoot button on the computer screen three times to capture three images. Then, manually select the collected images of the face and the tongue. Filter them according to the Image Acquisition page of the software.
Collect three images from each patient at a time and export the images to a folder for manual screening. Next, select a standard, fully exposed, well-illuminated, and clear images, the sample for algorithm training and testing. Delete the images that do not meet the criteria of a standard image.
The criteria for exclusion are incomplete tongue and face exposure and images that are too dark. To perform tongue image segmentation, open LabelMe and click on the Open button in the upper left corner of the interface. Select the folder containing the image to be segmented and open the photos.
Click the Create Polygon button to track the points. Track the tongue and lingual shapes, name them according to the selected areas, and save. On finishing all tracking, click Save to save the image to the data folder.
Set the device to capture images with a size of 1080x1920 pixels and ensure the size of the filled image is 1920x1920 pixels. Manually screen the the images of each patient to select and retain one from the three images that were taken for uncontrollable factors, such as subject blinking and lens blocking. For training the model, collect data from 200 individuals or 600 images and retain about 200 images that are usable after the screening.
Randomly divide all the tongue images according to the image number, placing 70%of them into the Training Set and 30%of the images into the Test Set folder. Download and install Anaconda Python and LabelMe from their official websites. Activate the environment and complete the installation and adjustment of the overall environment.
Choose the appropriate model for the research purpose. In this study, UNet, SegNet, DeepLab Version 3, and PSPNet were selected for validation. Next, build the deep learning algorithm model in the installed environment.
Tune the parameters by adjusting the values and complete the model training using the Training Set. Using LabelMe annotation and uniform image size methods, construct the required data set in conjunction with the research content. Examine the results of the segmentation and evaluate the model performance based on four metrics, precision, recall, mean pixel accuracy or MPA, and MIOU, to get a more comprehensive evaluation.
Compare the generated values of the four models. The higher the value, the higher the segmentation accuracy and the better the performance of the model. According to the index results, the UNet algorithm is superior to the other algorithms in MIOU, MPA, precision, and recall, and its segmentation accuracy is also higher.
PSPNet is better than DeepLab Version 3 in MIOU, MPA, and recall, while the DeepLab Version 3 model is lower than the segment model in all indexes. In tongue segmentation, the outer red area in the image results from tongue segmentation and the inner green area results from tongue coating segmentation. In this study, the unit algorithm performed well in tongue segmentation, and further research can be conducted based on the unit algorithms.
View the full transcript and gain access to thousands of scientific videos
Overview
This study utilizes deep learning algorithms, specifically U-Net, to segment tongue images effectively. The segmentation results are compared to evaluate the accuracy of tongue diagnosis.
Key Study Components
Area of Science
- Neuroscience
- Medical Imaging
- Deep Learning
Background
- Tongue diagnosis is a traditional method in Chinese medicine.
- Deep learning has shown promise in image segmentation tasks.
- Accurate segmentation can enhance diagnostic capabilities.
- Previous methods may lack precision in tongue image analysis.
Purpose of Study
- To segment tongue images using advanced deep learning techniques.
- To compare the performance of different segmentation algorithms.
- To identify the most effective algorithm for tongue image analysis.
Methods Used
- Utilization of U-Net and other deep learning models.
- Collection of tongue images using a handheld diagnostic instrument.
- Comparison of segmentation results from multiple algorithms.
- Assessment of segmentation clarity and accuracy.
Main Results
- U-Net demonstrated superior performance in segmenting tongue images.
- Clearer segmentation results were achieved compared to traditional methods.
- Algorithm performance varied, highlighting the need for careful selection.
- Results support the use of deep learning in medical diagnostics.
Conclusions
- Deep learning algorithms can significantly improve tongue image segmentation.
- U-Net is recommended for future tongue diagnosis studies.
- Further research is needed to refine these techniques.
Frequently Asked Questions
Executive Industry Relevance
Objectified tongue image segmentation using deep learning models addresses the need for standardized, reproducible diagnostic imaging in translational research. The comparative evaluation of segmentation algorithms, particularly U-Net, enhances predictive confidence in digital phenotyping and supports scalable data integration for biomarker discovery. Establishing robust image analysis workflows is critical for advancing digital pathology and cross-cohort data harmonization in biopharma R&D.
Strategic Applications in Biopharma R&D
Early Discovery & Target Validation
- Enables quantitative assessment of phenotypic features from standardized tongue images.
- Supports biological de-risking by providing objective, reproducible image segmentation outputs.
- Facilitates hypothesis-driven exploration of digital biomarkers in disease-relevant systems.
Screening & Assay Development
- Prepares validated image datasets for downstream algorithmic screening and model training.
- Improves assay reproducibility through uniform image acquisition and segmentation standards.
- Enables scalable evaluation of segmentation algorithms for platform reuse and screening readiness.
Translational & Preclinical Research
- Aligns digital image outputs with translational biomarker development initiatives.
- Supports continuity from image acquisition to preclinical validation of digital endpoints.
- Reduces ambiguity in phenotypic readouts, enhancing predictive value for translational studies.
Pipeline & Workflow Integration
This segmentation protocol integrates into the digital pathology workflow, spanning image acquisition, annotation, model training, and quantitative evaluation. It supports the transition from early discovery through preclinical research by enabling standardized, high-quality image datasets.
- Discovery Biology: Provides objective, quantitative image segmentation for hypothesis testing and pathway analysis.
- Screening: Delivers reproducible, validated image sets for algorithm benchmarking and assay development.
- Analytics: Generates precision, recall, and MIoU metrics to compare segmentation models and inform selection.
- Translational Research: Bridges digital image analysis with biomarker alignment and preclinical endpoint validation.
- Enterprise Reuse: Establishes a scalable, standardized workflow for future digital pathology and AI-driven studies.
Operational & Enterprise Impact
- Scientific Value: Increases predictive confidence and reduces mechanistic ambiguity in digital phenotyping.
- Operational Value: Standardizes image acquisition, annotation, and segmentation for reproducibility and scalability.
- Strategic Value: Enables robust go/no-go decisions and capital-efficient advancement of digital biomarker programs.
- Portfolio Impact: Supports risk-adjusted prioritization and cross-study data harmonization for enterprise R&D.
Implementation Considerations
- Requires expertise in deep learning, image annotation, and computational infrastructure.
- Depends on access to validated image acquisition instruments and annotation tools.
- Necessitates cross-team standardization of image quality and segmentation criteria.
- Must adapt protocols for diverse patient populations and imaging environments.
- Current limitations include lack of uniform standards and challenges with complex image backgrounds.
