Desarrollo editorial del género morisco en el romancero

Changes in relative concentrations of oxygenated, reduced, and total haemoglobin were estimated from DC PPG signals as done in conventional NIRS. Figure 9.4 illustrates an example of ∆HbO₂, ∆HHb, and ∆tHb estimated from PPG, along with the same pa-rameters estimated by NIRS. As can be seen from Figure 9.4, the relative haemoglobin concentrations estimated by PPG changed consistently during the occlusions and these responses are described in detail in this section.

Figure 9.5 shows the mean changes among all subjects of ∆HbO₂, ∆HHb, and ∆tHb estimated from PPG and NIRS during the venous occlusion in the first protocol. The occlusion at 60 mmHg occluded the veins, with a consequent congestion of venous blood in the forearm. The accumulation of venous blood can be seen in Figure 9.5 by the increase of reduced haemoglobin. This increase progressed until the occlusion was released after two minutes. The occlusion of the venous branch of the circulation in the forearm also caused an increase in concentration of the oxygenated haemoglobin, due to the blockage of the arterial-to-venous blood flow. The total haemoglobin tHb, which is the sum of oxygenated and reduced haemoglobin and represents the total blood volume, increased as well during venous occlusion. This corresponded to the incremental venous blood engorgement in the forearm. At the release of the occlusion, all parameters returned to baseline values.

9.3. Results

Figure 9.5: Mean (±SD) changes in relative haemoglobin concentrations estimated from PPG and NIRS during venous occlusion in the first protocol. Black traces are the means; grey traces are ± SD. ∆HbO2: oxygenated haemoglobin; ∆HHb: deoxygenated haemoglobin; ∆tHb: total haemoglobin. Vertical lines represent the inflation and deflation of the cuff. The occlusion pressure was 60 mmHg and it lasted for two minutes. PPG and NIRS measurements have been plotted on different y-axis scales to facilitate the visual assessment of the trends.

Figure 9.6 shows the mean changes among all subjects of ∆HbO₂, ∆HHb, and ∆tHb estimated from PPG and NIRS during the total occlusion in the first protocol. During total (over-systolic) occlusion, the reduced haemoglobin increased again due to the im-pediment of venous blood return. The occlusion of the arterial circulation caused a drop in oxygenated haemoglobin concentration. After slowly increasing at the beginning of the occlusion, the total haemoglobin ∆tHb reached a steady state. This was due to null arterial-to-venous flow caused by occlusion of both arterial and venous circulation in the forearm. The initial increase in ∆tHb at the commencement of the total occlusion was caused by a residual venous engorgement due to the non-instantaneous (i.e. 3-5 seconds) inflation of the cuff. At the release of the occlusion, the rapid increase of

∆HbO2 and ∆tHb indicated the post-occlusion reactive hyperaemia. After this, the signals returned to baseline readings.

Table 9.1 summarises the mean values of ∆HbO₂, ∆HHb, and ∆tHb estimated by both PPG and NIRS during the first protocol. The relative haemoglobin concentrations had significant changes from baseline (p < 0.05) in both venous and total occlusion.

9.3. Results

Figure 9.6: Mean (±SD) changes in relative haemoglobin concentrations estimated from PPG and NIRS dur-ing total occlusion in the first protocol. Black traces are the means; grey traces are ± SD. ∆HbO2: oxygenated haemoglobin; ∆HHb: deoxygenated haemoglobin; ∆tHb: total haemoglobin. Vertical lines represent the infla-tion and deflainfla-tion of the cuff. The occlusion pressure was 20 mmHg over the subjects’ systolic pressure and it lasted for two minutes. PPG and NIRS measurements have been plotted on different y-axis scales to facilitate the visual assessment of the trends.

Table 9.1: [Mean changes (±SD), in mM·cm, of relative haemoglobin concentrations changes estimated by PPG and NIRS during baseline, venous occlusion, and total occlusion. The values are the average of 30 s baseline, and the last 30 s of each occlusion.

Baseline Venous Occlusion Total Occlusion

∗ Significant change from baseline (p < 0.05)

9.3. Results

Figure 9.7: Mean (±SD) changes in relative haemoglobin concentrations during the second protocol. Black traces are the means; grey traces are ± SD. ∆HbO2: oxygenated haemoglobin; ∆HHb: deoxygenated haemoglobin; ∆tHb: total haemoglobin. Grey horizontal bars on top represent the duration of the occlu-sions. The order of the occlusions was 20 mmHg, 40 mmHg, 60 mmHg, 80 mmHg, 100 mmHg, subject’s systolic pressure, and over-systolic pressure. Each occlusion lasted for one minute and it was followed by one-minute recovery.

The second protocol aimed at investigating the responses to different degrees of vascular occlusions. Figure 9.7 shows the mean changes of ∆HbO₂, ∆HHb, and ∆tHb during the different occlusions of the second protocol. The relative haemoglobin concentrations estimated from PPG were able to respond to all the different occlusions induced. Even small venous occlusions of 20 mmHg and 40 mmHg were indicated by the parameters and the haemoglobin concentrations followed the responses of the same parameters measured by NIRS. Table 9.2 summarises the mean values of ∆HbO₂, ∆HHb, and

∆tHb estimated by the two methods during the steps of the second protocol. The table also shows the results of the significance tests, showing that there were significant changes from baseline (p < 0.05) in almost all occlusions.

The haemoglobin concentrations estimated from PPG presented some quantitative dif-ferences from the haemoglobin concentrations measured by NIRS. This is most probably due to the different penetration depths and the different wavelengths adopted. To inves-tigate this, the Bland & Altman plots were used to analyse these differences. However, since there cannot be limits of agreement set a priori, the ratio of the two measures

9.3. Results

able9.2:Meanchanges(±SD)in∆HbO2,∆HHb,and∆tHbestimatedbybothPPGandNIRSduringthesecondprotocol.ThevaluesareinmM·cmandwereaveragedover30sof andthelast30sofeachocclusion.Thereweresignificantchangesfrombaselineinalltheocclusions,exceptwhenindicated. OcclusionSteps(mmHg) Baseline20406080100SystolicTotal ∆HbO2(PPG)0.000±0.0240.040±0.0510.065±0.0600.061±0.0560.060±0.0570.024±0.048-0.026±0.0260.007±0.032∗ ∆HHb(PPG)-0.005±0.0080.013±0.0140.034±0.0180.030±0.0180.037±0.0160.030±0.0170.022±0.0140.021±0.013 ∆tHb(PPG)-0.006±0.0300.059±0.0610.099±0.0740.091±0.0660.091±0.0630.046±0.057-0.001±0.029∗ 0.024±0.036 ∆HbO2(NIRS)0.049±0.0750.115±0.0980.143±0.0970.137±0.1000.109±0.1030.062±0.1370.030±0.081∗ 0.052±0.111∗ ∆HHb(NIRS)0.005±0.0430.088±0.0650.102±0.0540.109±0.0490.108±0.0420.114±0.0640.083±0.0510.054±0.050 ∆tHb(NIRS)0.044±0.1040.186±0.1480.253±0.1430.255±0.1380.228±0.1430.190±0.1940.126±0.1070.120±0.124 ∗ Notsignificantchangefrombaseline(p>0.05)

9.3. Results

Figure 9.8: Bland & Altman plots and analysis of the haemoglobin concentrations estimated from PPG against the reference NIRS method during the first protocol. X-axes are the mean between the two methods, while the y-axes are the ratio of the measures. Blue crosses are the data. Horizontal solid lines are the standard mean bias and the dashed lines are the standard limits of agreement (mean ± 1.96SD). The black solid line is the regression line on the data, representing the mean bias. The grey lines are the ± 95 % confidence intervals.

The equation of the fitted regression line is reproduced at the bottom of the graphs. The original signals were resampled every 5 seconds.

was used instead of the absolute difference [185, 186]. This method still provides com-parative information on the two measures.

Figure 9.8 shows the Bland & Altman plots of the haemoglobin concentration esti-mated from PPG against the reference NIRS method during the first protocol. The dashed lines in the figure show the limits of agreements between the two methods, calculated with the standard Bland & Altman analysis [185]. The mean bias differ-ence ± 1.96SD (i.e. 95 % C.I.) for ∆HbO₂, ∆HHb, and ∆tHb were 0.9986 ± 0.0031, 0.9987 ± 0.0035, and 0.9973 ± 0.0039 respectively. These represent the mean bias ratio and suggest an overall underestimation of PPG against NIRS. However, the mean bias between the two methods varied over the ranges of measurements (average). In particular, the mean bias ratio between the measures seemed to decrease when the measurements increased. Thus, a regression line was fitted to the data as suggested by Bland and Altman [186]. The line and the 95 % C.I. are plotted on the data, along with the equation line.

The same Bland & Altman analysis was performed on the haemoglobin concentrations estimated during the second protocol. Figure 9.9 illustrates the Bland & Altman plots for the second protocol. The standard limits of agreement for ∆HbO₂, ∆HHb, and

∆tHb were respectively 0.9985 ± 0.0035, 0.9992 ± 0.0022, and 0.9978 ± 0.0047. These values were very similar to those calculated for the first protocol. In addition, the slopes of the fitted regression lines for the second protocol were very similar to the first protocol. This indicated the presence of a regular bias between the two techniques and

9.3. Results

Figure 9.9: Bland & Altman plots and analysis of the haemoglobin concentrations estimated from PPG against the reference NIRS method during the second protocol. X-axes are the mean between the two methods, while the y-axes are the ratio of the measures. Blue crosses are the data. Dashed lines are the standard mean and limits of agreement (mean ± 1.96SD). The black line is the regression line on the data, representing the mean bias. The grey lines are the ± 95 % confidence intervals. The equation of the fitted regression line is reproduced at the bottom of the graphs. The original signals were resampled every 5 seconds.

a good repeatability of the measurements.