Neural network study for standardizing pulse-taking depth by the width of artery.

Area of Science:

• Biomedical Engineering
• Traditional Chinese Medicine (TCM)
• Artificial Intelligence in Healthcare

Background:

• Traditional Chinese Medicine (TCM) pulse diagnosis relies on subjective depth definitions (Fu, Zhong, Chen).
• Existing depth definitions lack clarity for modern Pulse Diagnosis Instruments (PDIs).
• Need for a quantitative method to standardize pulse-taking depths.

Purpose of the Study:

• To propose a quantitative method for defining pulse-taking depths based on artery width (WA).
• To develop an index (α) for estimating WA for PDI applications.
• To improve the accuracy and reliability of pulse signal acquisition using PDIs.

Main Methods:

• Developed an artificial neural network (ANN) model using contact pressure (CP) and sensor displacement (SD) as inputs.
• Utilized ultrasound data for artery width (WA) and PDI data for CP and SD.
• Analyzed WA and SD data to train and validate the ANN model.

Main Results:

• The ANN model achieved a mean prediction error of 1.19% and a standard deviation (STD) of 0.0467.
• Significant improvement over the SD model, with a 71.62% reduction in mean prediction error and 29.78% in STD.
• The developed α value demonstrated reduced individual variation in estimating WA.

Conclusions:

• The proposed quantitative method and α index provide a reliable way to map artery width for PDIs.
• This approach standardizes pulse-taking depths, enhancing the consistency of TCM pulse diagnosis.
• Offers TCM physicians more dependable pulse information for clinical practice.