[Recognition system of acupuncture manipulations based on an array PVDF tactile sensor and machine learning]

Area of Science:

• Biomedical Engineering
• Acupuncture Research
• Sensor Technology

Background:

• Current acupuncture practice lacks objective physical parameters for manipulation techniques.
• Standardization and quantitative evaluation of acupuncture are crucial for education and clinical dissemination.
• Existing methods for assessing acupuncture manipulation are often subjective and lack precision.

Purpose of the Study:

• To design and develop a quantitative assessment tool for acupuncture manipulations.
• To enable objective measurement of exertion strength, duration, and direction in acupuncture.
• To create a system for the identification and classification of distinct acupuncture manipulation techniques.

Main Methods:

• Designed an array polyvinylidene fluoride (PVDF) tactile sensor to capture tactile parameters during needling.
• Developed a window segmentation method for tactile signal processing and extracted time-domain features.
• Constructed an acupuncture manipulation identification method utilizing Fuzzy C-Means (FCM) clustering.

Main Results:

• The PVDF tactile sensor effectively captured quantitative data during acupuncture manipulations.
• The proposed signal processing and FCM-based method successfully identified four basic acupuncture manipulations.
• Demonstrated the ability to distinguish between reinforcing and reducing techniques for both twirling/rotating and lifting/thrusting methods.

Conclusions:

• The developed PVDF tactile sensor system provides a quantitative basis for evaluating acupuncture manipulations.
• This technology facilitates the objective quantification and wider dissemination of acupuncture techniques.
• The findings support the integration of sensor technology into acupuncture education and practice for improved standardization.