神经网络在针灸训练系统中的应用研究

Application Research of Neural Network In Acupuncture Training System

针灸是中国特有的一种治疗疾病的疗法,但现有针灸训练方法的效率太低。提出了一种新型的针灸训练系统,建立更接近于针灸临床实际场景的针灸手臂硅胶模型,并在模型的相应穴位内置高精度压力传感器。然后以此模型为硬件依托,利用上位机软件来提取针刺力并分析所用的针灸手法。对获得的针刺力数据,应用神经网络来对其进行训练和分析,从而对针灸手法进行判别,以此来规范针灸学员的施针手法,实验结果表明神经网络算法判别准确率在93%-98%之间,可广泛应用于针灸训练系统。

Acupuncture is a unique treatment for diseases in China.However,the training efficiency of the existing acupuncture training method is too low.A new acupuncture training system is proposed to make a silicone model of acupuncture arm which is closer to the actual clinical scenes of acupuncture,and the high precision pressure sensors are built into the corresponding acupoints of the model.Based on this model,the PC software is used to extract the acupuncture force and to judge the acupuncture manipulation.The neural network algorithm is used to train and analyze the acupuncture force data,so as to judge the acupuncture manipulation,and to standardize the acupuncture manipulation of acupuncture trainees.The experimental results show that the accuracy of the neural net work algorithm is 93%-98%,which can be widely used in the acupuncture training system.