Chapter 2. Materials and Methods
2.2 Experimental instruments and procedure
Figure 5 is the experimental environment. In order to having stabilization for subjects and indoor environment, the experiment was conducted in an Electromagnetic Interference (EMI) shielding room (Acoustic Systems, Inc., Austin, USA) isolating subjects without any external stimulus and eliminating unexpected signal interference and electrical noise (shielding effectiveness per IEEE-299: 100 dB, 10 kHz–40 GHz). Structure of room (2.13 m long × 2.74 m wide × 2.42 m high) is sealed off completely, and just has single air outlet to control airflow passing. The hermetic character and single air outlet can be treated as a poor ventilation environment under fixed indoor gas exchange rate. Although some research indicated that indoor ventilation rates which is up to about 25 L s-1 per person, is associated with reduced prevalence of sick building syndrome (SBS) [20], but the room ventilation rate is controlled to be lower than 12.5 L s-1 for simulating a poor ventilation environment.Indoor Ta and CCO2 were monitored by a commercial hand-held computerized CO2 monitor (ZG-106;
Radiant Innovation, Inc., Hsinchu, Taiwan) and located at air entrance, air exit, and indoor instrument placement. The Tf was captured non-contactly by an infrared camera (IR FlexCam;
Everett, USA). The ECG and SpO2 were acquired simultaneously by ECG recorder and digital pulse oximeter respectively (OSTAR Meditech Corp., Taipei, Taiwan).
Fig. 5 Monitoring environmental conditions, three commercial hand-held computerized CCO2 monitors were used.
Monitor I, II, and III were located at air exit, indoor instrument placement, and air entrance respectively.
9
This thesis was submitted to the Institutional Review Board, National Taiwan University Hospital, Hsinchu Branch. To investigate the interaction between IAQ index (especially Ta and CCO2) and humans, the experimental procedure was divided into two phases for long-term environmental monitoring and humanity experiment. At first phase, indoor Ta and indoor CCO2 were monitored every 7 seconds without/with humans during 12- and 2-hour experiment periods. The purpose of 14-hour environmental monitoring is to discriminate the variation of environmental conditions without/with human, and to become the references for humanity experiment.
Fig. 6 Two experimental procedures. The first managed the environmental index for 14 hours in an airtight room 251.5 cm long × 191 cm wide × 196 cm high. The second procedure monitored physiological signals of subjects for 1 hour.
The second phase was to simultaneously record physiological signals. In total, 6 males and 4 females healthy volunteers were used in this phase. All subjects were selected randomly with no history or evidence of cardiac or pulmonary diseases, whose age ranged between 20 and 35 years (mean ± one standard deviation (SD) is 26±5 years), height was 168.2±6.8 cm, and weight was 64.0±16.2 kg. Before experiment, subjects gave written informed consent for the experimental protocol, which included an individual physiological baseline test. The experiment implemented 1 hour in afternoon at about 14:00-16:00. One subject stays in this room and plays puzzle game per trial for 1 hour in each trial. The Ta, CCO2, electrocardiography (ECG), SpO2, and Tf were acquired continuously during each trial.
10 Fig. 7 The user interface of the recording system
2. 3 Analysis methods
The analysis methods which include modeling, ANOVA, and t-test were applied to examine the relationship between different CCO2 and human physiological responses. To investigate how the different CCO2 affect in cardiovascular system, R peaks to R peaks interval (RRI) were calculated from ECG signals. All values are given as mean ± SD. Comparison between groups was analyzed using a one-way ANOVA to analyze RRI variation from low (600-800 ppm) to high (800-1200 ppm) CCO2, selecting half of time in the experimental period to separate low and high CCO2 in this stage. This study also used one sample t-test to examine whether the means of two groups are statistically different from each other. Significance was established at P < 0.05. According to the variation of SpO2 depends on CCO2 changes over time, SpO2 observes to different CCO2 and estimated to 0.01%. The variations of Tf alsobe concerned, the regression method which denotes as yˆi f(xi) can find the trend about the interaction between CCO2 andTf.
2. 3. 1 Statistic analysis
One-way ANOVA is a method of testing the equality of three or more population means
11
by analyzing sample variances. It is the method for testing hypothesis that three or more population means are equal, so a typical null hypothesis will be H0 and the alternative hypothesis (H1) will be the statement that at least one means is different from the others.
Based on sample means, ANOVA is also one of method to discuss the variance between and within samples.
at least one means is different
The ANOVA method is based on F distribution, which is a continuous probability
distribution. The F distribution arises frequently as the null distribution of a test statistic, most notably in the ANOVA. Although the details of the calculations are complicated, the easy way to interpret results is based on P-value. If P-value is small, such as 0.05 or lower, reject equality of means. Otherwise, fail to reject equality of meas.
Fig. 8 Relationship between F test and P-value
12
2. 3. 2 Curve fitting & Correlation
The main object of curve fitting is to describe the association between CCO2 samples by finding the graph and equation of the straight line that represents the association. This straight line is called the regression line (or best fit, or least-squares line), and its equation is called the regression equation. The simple regression equation is show as equation (2) [23].
(1) Where b0 is y-intercept and b1 is slope.
Regression equations are often useful for predicting and modeling the value of one variable, given some particular value of the other variable. If the regression line fits the data well, then it makes sense to use its equation for predictions, provided that we don’t go beyond the scope of the available values [23].
The correlation is to analyze a collection of paired sample data and determine whether there appears to be an association between the two variables. A correlation exists between two variables when one of them is related to the other in some way. The linear correlation coefficient R measures the strength of the linear association between the paired x- and y-quantitative values in a sample. Its value is computed by using equation (1). The linear correlation coefficient is sometimes referred to as Pearson product moment correlation coefficient in honor of Karl Pearson (1857-1936), who originally developed it.
(2) Where n represents the number of pairs of data present, x is x-values, y is y-values, and R is
13
the linear correlation coefficient for a sample. The value of R is always between -1 and +1 inclusive. Additionally, the value of R2 is the proportion of the variation in y that is explained by the linear association between x and y.
This research used conception of regression to model environmental condition, and used correlation coefficient to verify the accuracy between samples and modeling regression. The following equation demonstrates the correlation coefficient (R) between samples and modeling regression,
(3)
where y is samples value, and i yˆ is the modeling value. i