Cao Yujie, Li Ping, Zhu Yirun, Wang Zheng, Tang Nuo, Li Zhibin, Cheng Bin, Wang Fengxia, Chen Tao, Sun Lining
Jiangsu Provincial Key Laboratory of Advanced Robotics, School of Mechanical and Electric Engineering, Soochow University, Suzhou 215137, China.
Shanghai University of Medicine & Health Sciences, Shanghai 201318, China.
ACS Sens. 2025 Jan 24;10(1):272-282. doi: 10.1021/acssensors.4c02395. Epub 2025 Jan 6.
Atrial fibrillation (AF) as one of the most common cardiovascular diseases has attracted great attention due to its high disability and mortality rate. Thus, a timely and effective recognition method for AF is of great importance for diagnosing and preventing it. Herein, we proposed a novel intelligent sensing and recognition system for AF which combined Traditional Chinese Medicine (TCM), flexible wearable electronic devices, and artificial intelligence. Experiment and simulation synergistically verified that the flexible pressure sensor arrays designed according to the TCM theory could synchronously obtain the 3D pulses at Cun, Guan, and Chi. Combined with a homemade signal acquisition system and the pulse signals labeled by doctors of cardiovascular diseases, the differences in the 3D pulse signals between ones with AF and without can be picked up clearly. Enabled the convolutional neural network (CNN) and the pulse database, the recognition model was formed with a recognition rate of up to 90%. As a proof of concept, the artificial intelligence-enabled novel atrial fibrillation diagnosis system has been used to detect patients with AF in hospitals, showing 80% recognition rate. This work provides a new strategy to precisely diagnose and remotely treat AF, as well as to accelerate the development of Modern Chinese Medicine treatment.
心房颤动（AF）作为最常见的心血管疾病之一，因其高致残率和死亡率而备受关注。因此，一种及时有效的房颤识别方法对于其诊断和预防至关重要。在此，我们提出了一种新颖的房颤智能传感与识别系统，该系统结合了中医（TCM）、柔性可穿戴电子设备和人工智能。实验和模拟协同验证了根据中医理论设计的柔性压力传感器阵列能够同步获取寸、关、尺处的三维脉象。结合自制的信号采集系统和心血管疾病医生标注的脉象信号，可以清晰地捕捉到房颤患者和非房颤患者三维脉象信号的差异。利用卷积神经网络（CNN）和脉象数据库，形成了识别率高达90%的识别模型。作为概念验证，基于人工智能的新型房颤诊断系统已用于医院对房颤患者的检测，显示出80%的识别率。这项工作为精确诊断和远程治疗房颤以及加速现代中医治疗的发展提供了新策略。
