**3. Application to children**

The compact F-SAS sensor is made portable by it having only 15.3% of the volume of a conventional F-SAS sensor, as is shown in **Figure 20**. Under the approval of the Ethics Committee at the Dept. of Pediatrics at the University of Yamanashi, the subject wore a portable sleep apnea syndrome examination apparatus (SAS 2100) while a plastic optical fiber sheet was placed under the subject's bed. The subject's apnea/hypopnea index (AHI) was measured by the SAS 2100 and the respiratory disturbance index (RDI) was measured with new analysis software developed for the compact F-SAS sensor.

Further, this type of F-SAS sensor and Alice PDX of Philips-Respironics GK, a PSG device used to inspect sleep for OSAS diagnosis, were used for measurement. **Figure 21** below shows a diagram of the F-SAS sensor and PSG (Alice PDX).

#### **3.1 Results**

Thirty-eight of the 42 examinees show Pro-AHI > 5 and were suspected of SAS. The coincident measurement of ODI3% and Pro-AHI shows significant correlation (r = 0.79, p < 0.01). Seventeen examinees with Pro-AHI were over 10 and were eventually given SAS outpatient consultation. Four of them received a complete checkup by PSG and were given CPAP (Continuous Positive Airway Pressure) therapy. Next, a comparison of the analytical data of RDI (Pro-AHI) by the compact F-SAS sensor system and the coincident measurement of ODI3% by PULSOX are shown in **Figure 18**. Out of 76 examinees, 56 were Pro-AHI > 5 and were suspected of SAS. The coincident measurement of PULSOX, Pro-AHI, and ODI3% shows good correlation (r = 0.796, p < 0.01). The 32 examinees were consulted during SAS outpatient screening. In this manner, both the conventional type and the compact

*Sino-Nasal and Olfactory System Disorders*

**Figure 18.**

**Figure 19.**

**146**

*Correlation between the portable F-SAS sensor's pro-AHI and ODI3% of PULSOX.*

*Correlation between the portable F-SAS sensor's pro-AHI and AHI of PSG.*

Their correlation value was then obtained. The new analysis software was made up of a new algorithm that is difficult to use when the number of body movements per hour exceeds a certain number, and it analyzes the RDI by checking both the gasping judgment and weakness judgment against the data from a seriously ill patient. In this study, we improved the algorithm for severe markers in children (2–12 years old) and aimed to expand the application area to the compact F-SAS sensor. We compared severe markers obtained with the compact F-SAS sensor and existing simple polysomnography (PSG) (SAS 2100 manufactured by Nihon Kohden).

In this study, 27 data collected by the compact F-SAS sensor were analyzed by using the new analysis software, which were a severe marker caused by RDI, severe markers with periodic distribution, severe markers due to gasping respiratory

**Figure 20.**

*Comparison of compact F-SAS sensor and conventional one (left) and SAS 2100 (right).*

**Figure 21.** *F-SAS sensor (left) and PSG (right).*

frequency, severe markers due to hypopnea, and severe marker due to body movement frequency. The results are shown below. Before introducing the markers, the correlation value of R was 0.697 as shown in **Figure 22**.

After introducing the severe markers obtained by RDI, R = 0.58 as shown in **Figure 23***.* RDI 10 or more was considered as a serious patient and examined.

After introducing the severe markers due to gasping respiratory frequency, R = 0.76 as shown in **Figure 24**. We divided the total number of detected low respiration times by total landing time, and assumed that one with a high number of low respirations per hour is a serious one.

**Figure 23.**

**Figure 24.**

*hypopnea.*

**Figure 25.**

*frequency.*

**149**

*more was considered as a serious patient and examined.*

*Optical Fiber-Based Sleep Apnea Syndrome Sensor DOI: http://dx.doi.org/10.5772/intechopen.91060*

*Scatter diagram of F-SAS RDI/AHGI and SAS 2100 AHI after introduction of severe marker that is RDI 10 or*

*Scatter diagram of F-SAS RDI/AHGI and SAS 2100 AHI of severe marker algorithm with the number of*

*Scatter diagram of F-SAS RDI/AHGI and SAS 2100 AHI of severe marker algorithm based on body movement*

After introducing the severe markers due to hypopnea, and R = 0.87 after introducing the severe markers due to body movement frequency, as shown in **Figure 25**.

**Figure 22.**

*Scatter diagram of F-SAS RDI and SAS 2100 AHI of severe marker algorithm by RDI before introduction of severe marker.*

*Optical Fiber-Based Sleep Apnea Syndrome Sensor DOI: http://dx.doi.org/10.5772/intechopen.91060*

**Figure 23.**

frequency, severe markers due to hypopnea, and severe marker due to body movement frequency. The results are shown below. Before introducing the markers, the correlation value of R was 0.697 as shown in **Figure 22**.

low respirations per hour is a serious one.

**Figure 21.**

**Figure 22.**

**148**

*severe marker.*

*F-SAS sensor (left) and PSG (right).*

*Sino-Nasal and Olfactory System Disorders*

After introducing the severe markers obtained by RDI, R = 0.58 as shown in **Figure 23***.* RDI 10 or more was considered as a serious patient and examined. After introducing the severe markers due to gasping respiratory frequency, R = 0.76 as shown in **Figure 24**. We divided the total number of detected low respiration times by total landing time, and assumed that one with a high number of

After introducing the severe markers due to hypopnea, and R = 0.87 after introducing the severe markers due to body movement frequency, as shown in **Figure 25**.

*Scatter diagram of F-SAS RDI and SAS 2100 AHI of severe marker algorithm by RDI before introduction of*

*Scatter diagram of F-SAS RDI/AHGI and SAS 2100 AHI after introduction of severe marker that is RDI 10 or more was considered as a serious patient and examined.*

**Figure 24.**

*Scatter diagram of F-SAS RDI/AHGI and SAS 2100 AHI of severe marker algorithm with the number of hypopnea.*

**Figure 25.**

*Scatter diagram of F-SAS RDI/AHGI and SAS 2100 AHI of severe marker algorithm based on body movement frequency.*

We divided the total number of detected body movements by total bedtime, and assumed that one with more body movements per hour is a serious one.

Finally, we summarized as shown in **Table 2**. Sever markers were as *RDI*, *Periodic distribution*, *Candle breathing frequency*, *Low breathing frequency* and *Number of body movements*. As a result, *Number of body movements* showed best data of correlation value is 0.87.

The analysis results revealed the correlation value to be R = 0.87, which is a significant improvement over the correlation value of R = 0.697 between AHI obtained by SAS 2100 and RDI obtained by the conventional F-SAS sensor. From this, the software using the new algorithm effectively analyzes the RDI. In the future, to further improve the sensitivity of the compact F-SAS sensor, the remaining gap between the AHI and RDI in seriously ill patients must be bridged. To do this, algorithms will need to be developed that allow the analysis software of F-SAS sensors to capture highly disturbed respiration in seriously ill patients.

Medicine (AASM) in 2007, apnea events were judged from the waveforms of PSG.

son time was 21:18:39 to 23:36:24. His height was 99.1 cm, and his weight was 16.6 kg. The respiration cycle was 6 seconds/turn. Concerning the maximum amplitude of the thermal sensor, which was 24 (unit unknown), a drop of 90% or more can be seen for 15 seconds. **Table 3** shows that all five apnea event durations of F-SAS and PSG were consistent. As a result of comparing apneic events between F-SAS sensor and PSG, five apneic events in five out of five matched. In this

subject, all apnea events could be detected by the F-SAS sensor.

The subject was a boy, aged 4 years and 0 months. The measurement compari-

Clinical application of the F-SAS sensor showed that the respiratory disturbance index (RDI) from the F-SAS sensor corresponded well with the apnea-hypopnea index (AHI) from polysomnography under the AASM criteria published in 2001. The concurrently measured RDI and AHI had a correlation coefficient of 0.71. This means that the F-SAS sensor is well suited for preliminary SAS screening. The sensor would also be useful for screening potential SAS sufferers during normal sleep at home and for SAS screening of and monitoring of the respiration and heartbeat of neonates. Also we demonstrated through a detailed examination by PLSX and PSG (Alice5) that PLSX data and PSG (Alice5) data are well correlated with those of the F-SAS sensor. The F-SAS sensor is effective for SAS screening

Further, a good correlation value of 0.87 was obtained between the RDI calcu-

This research was supported by the Japanese Ministry of Education, Science, Sports and Culture, Grant-in-Aid for Scientific Research, 19,656,101, in fiscal year 2007–2008 and partially supported by the Japan Science and Technology Agency, Grant-in-Aid for A-STEP High-risk Challenge, JST (A-STEP, #AS2114072A), in fiscal year 2009–2011. For the portable prototype manufacturing, I am grateful for

lated by a new compact F-SAS sensor analysis system using an algorithm for severely ill children and AHI calculated by using the SAS 2100. Also, by comparing the sleep-event results of the F-SAS sensor with those of the PSG, it was observed that apnea sleep events matched. This suggests that it might be valid applied to diagnosing sleep apnea with the pediatric F-SAS sensor. We further analyze the

The analysis results are as shown in **Figure 26**.

*Optical Fiber-Based Sleep Apnea Syndrome Sensor DOI: http://dx.doi.org/10.5772/intechopen.91060*

*Results of apnea events comparison between F-SAS sensor and PSG.*

during a full overnight medical check-up.

subject data for further improvement.

**Acknowledgements**

**151**

**4. Conclusions**

**Table 3.**

We improve the accuracy of the software used for analysis with the F-SAS sensor for child development. This was done for the development of PSG equipment of Philips Respironics GK used in a pediatric clinical trial at Yamanashi University School of Medicine for the purpose of comparing apneic events between the pediatric F-SAS sensor and PSG. That is, the respiratory events were judged and compared from the waveform data at bedtime, which is the analysis result of the F-SAS sensor and PSG. On the basis of the definition of the scoring rule of children for sleeping respiratory events, proposed by the American Academy of Sleep


*Correlation value in each severe marker.*

**Figure 26.** *Measurement analysis result of PSG (apnea waveform).*

*Optical Fiber-Based Sleep Apnea Syndrome Sensor DOI: http://dx.doi.org/10.5772/intechopen.91060*


#### **Table 3.**

We divided the total number of detected body movements by total bedtime, and

Finally, we summarized as shown in **Table 2**. Sever markers were as *RDI*, *Periodic distribution*, *Candle breathing frequency*, *Low breathing frequency* and *Number of body movements*. As a result, *Number of body movements* showed best data of

The analysis results revealed the correlation value to be R = 0.87, which is a significant improvement over the correlation value of R = 0.697 between AHI obtained by SAS 2100 and RDI obtained by the conventional F-SAS sensor. From this, the software using the new algorithm effectively analyzes the RDI. In the future, to further improve the sensitivity of the compact F-SAS sensor, the remaining gap between the AHI and RDI in seriously ill patients must be bridged. To do this, algorithms will need to be developed that allow the analysis software of F-SAS sensors to capture highly disturbed respiration in seriously ill patients. We improve the accuracy of the software used for analysis with the F-SAS sensor for child development. This was done for the development of PSG equipment of Philips Respironics GK used in a pediatric clinical trial at Yamanashi University School of Medicine for the purpose of comparing apneic events between the pediatric F-SAS sensor and PSG. That is, the respiratory events were judged and compared from the waveform data at bedtime, which is the analysis result of the F-SAS sensor and PSG. On the basis of the definition of the scoring rule of children for sleeping respiratory events, proposed by the American Academy of Sleep

assumed that one with more body movements per hour is a serious one.

correlation value is 0.87.

*Sino-Nasal and Olfactory System Disorders*

**Table 2.**

**Figure 26.**

**150**

*Correlation value in each severe marker.*

*Measurement analysis result of PSG (apnea waveform).*

*Results of apnea events comparison between F-SAS sensor and PSG.*

Medicine (AASM) in 2007, apnea events were judged from the waveforms of PSG. The analysis results are as shown in **Figure 26**.

The subject was a boy, aged 4 years and 0 months. The measurement comparison time was 21:18:39 to 23:36:24. His height was 99.1 cm, and his weight was 16.6 kg. The respiration cycle was 6 seconds/turn. Concerning the maximum amplitude of the thermal sensor, which was 24 (unit unknown), a drop of 90% or more can be seen for 15 seconds. **Table 3** shows that all five apnea event durations of F-SAS and PSG were consistent. As a result of comparing apneic events between F-SAS sensor and PSG, five apneic events in five out of five matched. In this subject, all apnea events could be detected by the F-SAS sensor.
