EEXPERIMENTAL RESULTS AND DISCUSSION
In order to evaluate the performance of the proposed sleep
monitoring system, 7 healthy subjects (5 male, 2 female)
were recruited for lab-controlled experiments. A biosignalsplux sensor node [25], which includes a 3-axis accelerometer,
was placed onto the chest, as shown in the Fig. 5. For
comparison, the ECG signal was also recorded by the sensor
node for heart rate extraction as the ground-truth. Meanwhile,
the pulmonary gas exchange measurement (V02) was recorded
by the K4b2 system [26] for the respiratory rate extraction.
Once the acceleration is collected, it will be analysed off-line
using the proposed method.
A. sleep posinoti identification results
The acceleration signal was firstly applied to sleep posture
recognition when static postures are detected. Fig. 6 shows
the confusion matrices to illustrate the class-to-class average
performance for all the subjects using the LDA classier. More
EEXPERIMENTAL RESULTS AND DISCUSSIONIn order to evaluate the performance of the proposed sleepmonitoring system, 7 healthy subjects (5 male, 2 female)were recruited for lab-controlled experiments. A biosignalsplux sensor node [25], which includes a 3-axis accelerometer,was placed onto the chest, as shown in the Fig. 5. Forcomparison, the ECG signal was also recorded by the sensornode for heart rate extraction as the ground-truth. Meanwhile,the pulmonary gas exchange measurement (V02) was recordedby the K4b2 system [26] for the respiratory rate extraction.Once the acceleration is collected, it will be analysed off-lineusing the proposed method.A. sleep posinoti identification resultsThe acceleration signal was firstly applied to sleep posturerecognition when static postures are detected. Fig. 6 showsthe confusion matrices to illustrate the class-to-class averageperformance for all the subjects using the LDA classier. More
การแปล กรุณารอสักครู่..
