Fig 1.
Illustration of the experimental setup including RIP thoracic belt, nasal cannula and a polygraph for acquisition of data.
Fig 2.
Description of the custom made algorithm in 3 steps.
Step 1: recognition of the onset of the respiratory cycle by nasal signal (vertical bars). Step 2: searching for local RIP signal maximums (dark squares). Step 3: searching for local RIP signal minimums (light squares). Illustration of the results after treatment by the algorithm 1(bottom panel): treated signal (dotted line) superimposed on the raw signal (continuous line).
Table 1.
Summary of statistical analysis comparing the RIP signal processed by the custom made algorithm with PT.
r: Spearman coefficient of correlation, p: p-value from Spearman coefficient determination, CI: Confidence intervals, SD: Standard deviation.
Fig 3.
Evaluation of Vt, Ti and Te by RIP signals processed by the custom made algorithm.
The linear relationship between Vt determined by PT, and Vt determined by RIP plus the algorithm is shown in the upper left panel. Bland and Altman’s analysis of Vt determined by PT, and Vt determined by RIP plus the algorithm is shown in the upper right panel with bias (long dotted line) and limit of agreements (short dotted line). The linear relationship between Ti determined by PT, and Ti determined by RIP plus the algorithm is shown in the middle left panel. Bland and Altman’s analysis of Ti by PT and by RIP plus the algorithm is shown in the middle right panel with bias (long dotted line) and limit of agreements (short dotted line). The linear relationship between Te determined by PT, and Te determined by RIP plus the algorithm (lower left panel). Bland and Altman’s analysis of Te by PT and Te by RIP signal plus the algorithm (lower right panel) with bias (long dotted line) and limit of agreements (short dotted line).
Fig 4.
Evaluation of Ti/Ttot and Vt/Ti ratios determined by RIP signals processed by the custom made algorithm.
The linear relationship between Ti/Ttot ratio by PT and the Ti/Ttot ratio by RIP signal treated by the algorithm (upper left panel). Bland and Altman’s analysis of Ti/Ttot ratio determined by PT and Ti/Ttot ratio determined by RIP signal treated by the algorithm (upper right panel) with bias (long dotted line) and limit of agreements (short dotted line). Linear relationship between Vt/Ti ratio determined by PT and the Vt/Ti ratio determined by RIP signal treated by the algorithm (lower left panel). Bland and Altman’s analysis of Vt/Ti ratio determined by PT and Vt/Ti ratio determined by RIP signal treated by the algorithm (lower right panel) with bias (long dotted line) and limit of agreements (short dotted line).
Fig 5.
Evaluation of slopes from linear relationships between RIP and PT before and after activity.
Variations in slopes (left panel) ns: not significant. Relationship between slopes before and after activity (right panel) r: Spearman correlation coefficient; p: p-value.