**3. Physiological data presentation**

30 Health Management – Different Approaches and Solutions

Fig. 12. Interactions among smart phone, Reader, and Tag

Fig. 13. Flow chart of sending SMS text message under emergency situation

The hardware is implemented as Fig. 6. The Bluetooth adaptor is connected to the RFID reader and the ring Tag is worn on the user's finger. The setup steps for communication between the smart phone and the RFID reader/Tag can be executed either on the smart phone emulator (on PC) or on the smart phone using the Cellular Emulator. Fig. 14 illustrates these steps on the smart phone emulator.

Fig. 14. Setup steps for communication between the smart phone and RFID reader/tag

A Mobile-Phone-Based Health Management System 33

Fig. 16. GUI for emergency numbers

Fig. 17. GPS information


Fig. 15 depicts the physiological data monitoring GUI on the smart phone emulator. The same GUI on the smart phone is shown in Fig. 10. On this page, two temperature values measured by RFID reader and Tag and pulse data are shown. The SOS message box indicates the status of the SOS button on the ring Tag. If this button is pressed, the smart phone will dial the pre-set emergency phone number(s) to send out SMS with the GPS position information to other people for help. Fig. 16 shows the GUI for setting up pre-set emergency phone number(s). Fig. 17 shows the GPS information.

Fig. 15. Physiological data monitoring GUI

**Step 1.** On the main page, click the "Menu," go to "Setting," and then choose "RFID

**Step 2.** Choose the correct COM port and click "OPEN." Then, click "BACK" to return to

**Step 3.** Click "Reader" and "Reader setup" page will pop up. Click "ReaderReset" to reset

**Step 4.** Click Tag and the "Tag setup" page will pop up. First click "ALLReset" and turn on

Fig. 15 depicts the physiological data monitoring GUI on the smart phone emulator. The same GUI on the smart phone is shown in Fig. 10. On this page, two temperature values measured by RFID reader and Tag and pulse data are shown. The SOS message box indicates the status of the SOS button on the ring Tag. If this button is pressed, the smart phone will dial the pre-set emergency phone number(s) to send out SMS with the GPS position information to other people for help. Fig. 16 shows the GUI for setting up pre-set

the RFID reader, and the RFID reader beeps and "True" is shown on the message box. Then click "ReaderQuery" to search for the Reader's address. The message box

the power of the ring Tag. Then click "SearchTag" to search for the ring Tag around the RFID reader. Once the ring Tag is detected, the Tag ID number will be shown on the page and finally click "Access" to go to the physiological data monitoring

setting." RS232 COM port setup page will pop up.

of ReaderQuery shows the address number.

emergency phone number(s). Fig. 17 shows the GPS information.

Fig. 15. Physiological data monitoring GUI

the main page.

GUI.

Fig. 16. GUI for emergency numbers


Fig. 17. GPS information

A Mobile-Phone-Based Health Management System 35

estimation of sleep depth and physical activities. In Fig. 24, four indices listed below are

3. EEG power ratio of gamma band (30*~*128 Hz) to delta band (below 4 Hz) derived from

4. EEG power ratio of alpha band (8*~*13 Hz) to beta band (13*~*30 Hz) derived from 128-

These indices are calculated for every 3-minute periods, and time courses of them are shown. Note that the index value for a particular 3-minute period disappears if more than 5% of the raw waveform exceeds the A/D conversion range during the period, for example

From examination of Fig. 24 and subject's handwritten note, the availability of these indices for activity and sleep depth estimation can be pointed out. Both the median of heart rate and the standard deviation of accelerogram are low during sleep. Therefore these indices are useful for estimation of sleep/wake state. More complicated algorithm may enable us to estimate physical activities from these biosignals. The EEG power ratio of gamma band to delta band shows sharp peaks. The peak times coincide with mealtimes of subject's record. This EEG ratio surges when chewing EMG appears and is useful for estimation of mealtime. The EEG power ratio of alpha band to beta band changes periodically during sleep periods. This phenomenon suggests that this index may reflect changes in sleep depth. However, further investigation is required for more rigorous

shown for the waveforms during the 4-day measurement:

128-second EEG waveform,

Fig. 19. Waveforms during deep sleep

second EEG waveform.

conclusions.

1. Median value of heart rate derived from 3-minute ECG waveform, 2. Standard deviation for 3-minute absolute values of accelerogram,

in the EEG ratios at the first day's night (from about 20 to midnight).

Fig. 18 presents the 3G communication of smart phone and the webpage of remote medical station. The 3G connection setup page of the smart phone is shown on the left side and the webpage of the server is on the right side where temperature values, pulse, and Google map are displayed. Based on the GPS position information sent from the smart phone, the Google map helps the medical staff to locate promptly the monitored user who needs further assistance.

Fig. 18. 3G communication setup and the webpage of remote medical station

For evaluation of the developed biosignal recorder, a 4-day measurement was performed. Figures 19–21 show typical waveforms of EEG, ECG and accelerogram during the 4-day measurement. Fig. 19 represents waveforms during deep sleep. In the EEG waveform, sleep spindles appeared frequently. The heart rate was very slow (42 bpm) as seen in ECG waveform. The accelerogram did not change during this period. Fig. 20 was acquired during shallow sleep. The ECG and accelerogram are almost same to these in Fig. 19. On the other hand, sleep spindles are not seen in EEG waveform. In Fig. 21, waveforms during deskwork are shown. The heart rate (50 bpm) was faster than that during sleep. The accelerogram denotes that subject's body is upstand. In EEG waveform, large artifacts generated by eye blinks appear. Fig. 22 shows waveforms during walk movements. Up and down movements of the walking cycle are observed in the accelerogram. The heart rate was moderately fast (about 90 bpm) due to the walking movements. In Fig. 23, waveforms during meal are indicated. The EEG waveform was disturbed by chewing movements. This phenomenon was not desirable for true EEG recording, however, this characteristic waveform is useful for activity estimation The characteristics of these waveforms suggest some indices for

Fig. 18 presents the 3G communication of smart phone and the webpage of remote medical station. The 3G connection setup page of the smart phone is shown on the left side and the webpage of the server is on the right side where temperature values, pulse, and Google map are displayed. Based on the GPS position information sent from the smart phone, the Google map helps the medical staff to locate promptly the monitored user who needs further

GPS data

Physiologica

l

Google Map

Fig. 18. 3G communication setup and the webpage of remote medical station

For evaluation of the developed biosignal recorder, a 4-day measurement was performed. Figures 19–21 show typical waveforms of EEG, ECG and accelerogram during the 4-day measurement. Fig. 19 represents waveforms during deep sleep. In the EEG waveform, sleep spindles appeared frequently. The heart rate was very slow (42 bpm) as seen in ECG waveform. The accelerogram did not change during this period. Fig. 20 was acquired during shallow sleep. The ECG and accelerogram are almost same to these in Fig. 19. On the other hand, sleep spindles are not seen in EEG waveform. In Fig. 21, waveforms during deskwork are shown. The heart rate (50 bpm) was faster than that during sleep. The accelerogram denotes that subject's body is upstand. In EEG waveform, large artifacts generated by eye blinks appear. Fig. 22 shows waveforms during walk movements. Up and down movements of the walking cycle are observed in the accelerogram. The heart rate was moderately fast (about 90 bpm) due to the walking movements. In Fig. 23, waveforms during meal are indicated. The EEG waveform was disturbed by chewing movements. This phenomenon was not desirable for true EEG recording, however, this characteristic waveform is useful for activity estimation The characteristics of these waveforms suggest some indices for

assistance.

estimation of sleep depth and physical activities. In Fig. 24, four indices listed below are shown for the waveforms during the 4-day measurement:


These indices are calculated for every 3-minute periods, and time courses of them are shown. Note that the index value for a particular 3-minute period disappears if more than 5% of the raw waveform exceeds the A/D conversion range during the period, for example in the EEG ratios at the first day's night (from about 20 to midnight).

From examination of Fig. 24 and subject's handwritten note, the availability of these indices for activity and sleep depth estimation can be pointed out. Both the median of heart rate and the standard deviation of accelerogram are low during sleep. Therefore these indices are useful for estimation of sleep/wake state. More complicated algorithm may enable us to estimate physical activities from these biosignals. The EEG power ratio of gamma band to delta band shows sharp peaks. The peak times coincide with mealtimes of subject's record. This EEG ratio surges when chewing EMG appears and is useful for estimation of mealtime. The EEG power ratio of alpha band to beta band changes periodically during sleep periods. This phenomenon suggests that this index may reflect changes in sleep depth. However, further investigation is required for more rigorous conclusions.

Fig. 19. Waveforms during deep sleep

A Mobile-Phone-Based Health Management System 37

Fig. 22. Waveforms during walk

Fig. 23. Waveforms during meal

Fig. 20. Waveforms during shallow sleep

Fig. 21. Waveforms during deskwork

Fig. 20. Waveforms during shallow sleep

Fig. 21. Waveforms during deskwork

Fig. 22. Waveforms during walk

Fig. 23. Waveforms during meal

A Mobile-Phone-Based Health Management System 39

management system effective and convenient to help people on developing a healthy life

In this chapter, a mobile e-health-management system has been presented. This system integrates a wearable ring-type pulse monitoring sensor and a portable biosignal recorder with a smart phone. The ring-type pulse monitoring sensor can measure pulse and temperature, while the biosignal recorder can record electroencephalogram (EEG), electrocardiogram (ECG), and body 3-axis acceleration during daily lives. Based on these EEG, ECG, and acceleration data, several indices can be evaluated to estimate the user's physical activities. With the help of the mobile "exercise-333" health management mechanism developed on the smart phone, the user can monitor his/her own physiological data to practice the idea of "Prevention is better than cure," developing a healthier life style. The presented system, with the popularity and mobility of smart phones, effectively provides the needs for mobile health management.

The authors would like to express their gratitude to Sinopulsar Technology Inc., Taiwan for free use of their RFID ring pulse sensor in this work. Financial support from the National

style.

Fig. 25. Exercise-333 health management GUIs

Science Council, Taiwan are acknowledged as well.

**5. Conclusions** 

**6. Acknowledgement** 

Fig. 24. Indices for activity estimation
