New research bridges the distance between traditional Chinese medicine, modern technology
In Chinese medicine, one of the steps in determining the health of a person’s body is to examine the tongue. Dong Xu, chair of the computer science department, and Ye Duan, associate professor of computer science, and their research assistants, are developing an app that operates on this principle.
“We’d really like to develop an app that can be used at home that everyone can download from the app store,” Xu said.
The app will analyze images of the tongue based on an algorithm developed by Duan. It uses the color and coating of the tongue to classify the total health of the body, or zheng.
Zheng can be classified as hot or cold, but it does not refer solely to body temperature. Hot zheng is characterized by symptoms such as a preference for a cold environment, thirst, constipation, noticeable bad breath and acidic saliva. Chills characterize cold zheng more, as well as a preference for a warm environment and hot drinks or food, a high and short-pitched voice and pale flushing of the face.
“This traditional medicine has actually been around for about 5,000 years,” Xu said. “It’s accumulated a lot of useful empirical knowledge. Using information technology to utilize the value in traditional Chinese medicine and also bring more rigor and more precision to it — we think our work can probably help in that direction.”
Both hot and cold zheng can indicate gastritis, according to traditional Chinese medicine. Gastritis is an inflammation of the stomach lining and is sometimes caused by a bacterial infection.
Duan and Xu worked with Shao Li, a professor in the bioinformatics division at Tsinghua University in Beijing. Li approached Xu about the idea of adapting the traditional approach with modern technology.
“He actually collected the initial data that laid the foundation for the computational study we did here,” Xu said.
Duan used tongue images from 263 gastritis patients and 48 healthy individuals to develop the image analysis software. The images were classified by the intensity of their symptoms and whether they were infected with a certain bacteria. The gastritis patients had already been classified as having hot or cold zheng.
These images were used to train three supervised learning algorithms to diagnose the health of the patient based on the image.
“We used the machine learning technique with images from people diagnosed by doctors to teach the machine how to diagnose,” Duan said. “We had both training and testing data.”
The need for a large dataset to train the program and test its accuracy limited the research to focus on a single illness — gastritis. The software was able to correctly diagnose a gastritis patient 80 percent of the time, Duan said.
“The more data you have, the faster you learn — it’s just the same with the computer,” Duan said. “The key is to get more datasets.”
Xu agreed that more data would be necessary to enhance the usefulness of the app being developed.
“For basic research we need to collect a lot of data with one disease because if we try to have many diseases then likely we will only have a small amount of data for each,” Xu said. “From here we’d really like to have a more comprehensive analysis. It would take more data.”
The idea of an app to analyze health based on the appearance of the tongue has sparked widespread interest, with more than 100 websites covering this story. Xu said the response has been varied and noted that something so new usually concerns people. He emphasized that the app would encourage awareness of personal health.
“We really need to be cautious about how we describe this application. It’s not meant to replace clinical diagnosis but to be complimentary,” Xu said. “This is actually more like a warning tool. Often people don’t pay attention to their health status. They may feel some symptoms but not pay attention.”
- Computers & Electronics
- Health / Medicine
- Infrastructure & Transportation
- Nano Science & Technology
- National Security / Defense
- The Environment
- All Academic Departments
- Chemical Engineering
- Civil & Environmental Engineering
- Computer Science
- Electrical & Computer Engineering
- Industrial & Manufacturing Systems Engineering
- Information Technology
- Mechanical & Aerospace Engineering
- MU Informatics Institute
- Naval Sciences
- Nuclear Engineering Program
- Nuclear Science & Engineering Institute
- Back to menu
- Faculty & Staff
- Research Centers & Programs
- Mizzou Engineer Magazine
This story is tagged as:
- MU engineer lends water quality expertise to federal, state agencies
- MU researcher probing deep learning’s possibilities
- Silver nanoparticle concentration too low to be harmful in water supply, paper finds
- Dean selected as fellow of Biomedical Engineering Society
- Stem cell transformation research sheds new light on osteoporosis