| 摘要: | 透過心率偵測追蹤，對於保持健康或提升運動表現都是一項重要的生理指標。腕戴
式光學心率裝置利用光體積變化描記圖(PPG)偵測心率，不同於技術成熟且穩定的電極
式心電圖(ECG)裝置，PPG 裝置在各種不同的活動類型(休息、低//強度運動)有著
不一致的準確度分析。
本研究結果說明 PPG 裝置可作為偵測心率指標的良好工具，若排除各品牌的演算
法差異，單純探討 ECG/PPG 裝置之選擇與應用，考量 PPG 技術的進步及裝置配戴的方
便性，在生活休閒/運動健身/外探險等情境，相較於 ECG，PPG 裝置會是更好的選擇。
此項偵測心率的運動測試總共有三十二位參與者，運動測試組態包括靜息、騎室內自行
車的三種強度:功率 40 瓦/70 瓦/100 瓦、跑步機上行走(5 公里/小時)/跑(7 公里/小時)/
跑步(9 公里/小時)和肩上推舉等運動組態，而 Garmin Fenix 7s (PPG)腕戴式裝置與 HRM�
Tri (ECG)戴式裝置的心率數據呈現出顯著相關，全組態之相關係數 rc = 0.988，而區分
各組態觀察，靜息組態 rc = 0.915，室內自行車 rc = 0.983；跑步機組態 rc = 0.949，肩上
推舉組態 rc = 0.987，所有組態之顯著水準 p < 0.01；並將 Bland-Altman plot 結合心率區
間作細部探討，全組態心率區間以 zone 4 離散值最多，靜息組態只分佈在 zone 1，室
內自行車以 zone 4 離群值最多，步機組態以 zone 4 離群值最多，肩上推舉組態以
zone 1 離群值最多。;Heart rate(HR) detection is an important physiological indicator for maintaining health
and improving athletic performance. Wrist-worn optical heart rate devices utilize
photoplethysmography (PPG) technology to measure HR. However, compared to the well�
established and stable electrocardiogram (ECG) technology, wrist-worn optical HR detection
using PPG devices has shown inconsistent accuracy across different types of activities,
including rest, exercise of low/medium/high intensity.
The result showed that PPG devices can serve as effective tools for HR monitoring. When
considering the advancements in PPG technology and the convenience of device wearability,
PPG devices are a preferable choice for applications involving daily life, exercise & fitness,
and outdoor & adventures, compared to ECG devices, especially when excluding algorithm
differences among brands.
The exercise test for HR detection involved 32 participants and included various exercise
configurations, such as rest, three intensities of indoor cycling (40/70/100 watts), treadmill
walking (5 km/h)/jogging (7 km/h)/running (9 km/h), and push over. The HR data of the Garmin
Fenix 7s (PPG) wrist-worn device and the HRM-Tri (ECG) chest strap device show a
significant correlation. The overall correlation coefficient, rc is 0.988. When considering
different configurations, the correlation coefficients are as follows: resting configuration rc =
0.915; indoor cycling configuration rc = 0.983; treadmill configuration rc = 0.949; Push over
configuration rc = 0.987. The significance level for all configurations is p < 0.01.
Furthermore, the detailed analysis by combining the Bland-Altman plot with heart rate
zones. Among all configurations, zone 4 exhibits the highest number of outliers. In the resting
configuration, outliers are primarily distributed in zone 1. For the indoor cycling configuration,
zone 4 has the highest number of outliers. Similarly, the treadmill configuration shows the
highest number of outliers in zone 4, while the Push over configuration has the highest number
of outliers in zone 1. |
