Design and realization of a respiratory simulator to evaluate opto-electronic plethysmograph functions

Opto-electronics Systems (OS) are able to measure human movement without mechanical restraints, in order to describe kinematic and temporal features of the movement of various body segments. Actually, OS are used in different medical fields: gait analysis, human superior limb movement, respiratory pattern analysis and similar applications. Human movements are very complex, therefore quantitative and qualitative analysis may be useful. Different studies analyzed measurement's accuracy and precision of opto-electronic systems, in particular on liner displacements. Some clinical applications have needed to measure surface and volume of human body. Therefore, OS performance must be evaluated also for these particular measurements. The purpose of this work is to realize a volumetric respiratory simulator (RS), performing the thorax movements with known volume variations, in order to assess the reliability and accuracy of OS volume measurements in dynamic conditions. Opto-electronic plethysmograph (OEP) is the opto-electronic system considered in this work. The RS has been designed and realized in order to simulate quiet human breathing and to evaluate OEP's functions. OEP is employed in clinical applications to measure pulmonary volumes by starting markers positions. OEP allows to measure volume of Chest Wall (CW) and of six different compartments (right and left): Rib Cage Pulmonary (RCp), Rib Cage Abdominal (RCa) and Abdomen (AB). The number of marker changes in according to clinical application and position's patient: standing, supine/prone and seated. The triangulation is the process on which OS is based, to determine a point in 3D space given into two or more projections. First, stereo-triangulation model is implemented with MATLAB in order to obtain preliminary metrological characterization of OEP. The pixelation of CCD sensors is introduced in this model in order to evaluate the error due to CCD sensors discretization. Results show that the error increases by increasing the camera distance from the point. Then, preliminary accuracy evaluation on linear displacements has been carried out because volume measurements of OEP are based on marker's linear displacement. Linear displacements (ranged from 10 µm to 200 µm) were performed by a linear DC-motor and by varying systematically marker size (diameter 6 and 12 mm), number of cameras (2, 4, and 6) and displacements magnitude. Displacements of 10 µm are not discriminated by OEP, for any kind of marker and any number of cameras. Therefore, from further trials OEP showed a discrimination threshold of 30 µm. The largest errors are measured for displacement smaller than 50 µm. Afterwards, the respiratory simulator (RS) for OEP is designed to reproduce the thorax movements during normal human breathing. The RS volume must always be known in order to be compared with volume measured through OEP. Markers have to applied on the RS so that OEP is able to calculate the RS volume. At the same time, the RS volume has to be evaluated by another measurement system, such as the comparison is possible. In this way, OEP metrological characteristics can be identified. The RS is realized using linear actuators, placed on a fixed base. Actuators move 8 panels, on which 89 markers are placed. The fixed base is the supporting structure for each component of RS. Fixed panels (n°6) are bolted to fixed base and they hold the support motor. Actuators, rigidly coupled with mobile panels through screws, are placed on the fixed panel. Positions of 8 compartments and then of 89 marker placed on panels are known through Hall sensors, embedded in actuators. In this way, the RS volume can be estimated starting from 89 marker positions through an algorithm realized in Matlab. In order to verify OEP metrological characteristics on volume measurements, RS volume has been estimated both by OEP and by algorithm. Seven different trials have been carried out, setting the panels with different displacements. At first, RS volume has been calculated when all panels are motionless. In this way, static volume has been measured both through algorithm starting from panel initial positions, recorded by Hall sensors, and OEP. Results show that OEP overestimates the RS static volume of 0,1 L. Volumes measured by OEP have the same trend of volumes calculated by algorithm. Algorithm and OEP's volume have been compared for one act, in order to evaluate the correlation and R2. Then, average on 10 tidal volumes has been calculated for each displacement. Comparison between tidal volume measured by OEP and computed by algorithm show that measurements are not discrepant for small displacement (2 and 3.2 mm) and quiet breathing simulation (QB). Finally, the RS could be considered useful to analyse OEP tidal volumes only for small panel's displacements. That is not a problem for the final goal's achievement because chest wall movements do not exceed 5 mm during quiet breathing in normal subjects. The accuracy of OEP tidal volume measurements has been evaluated and it ranges from 8% to 18%. The best accuracy is obtained for 2 and 3.2 mm displacements. This result could be encouraging because usually the chest wall displacement is ranged from 2,5 mm to 4,5 mm. In conclusion, the RS appears being more suitable than spirometer, in order to validate OEP, overcoming issues as dynamic response and humidity, pressure and temperature's gas changes, introduced by spirometer.