Limits...
Enhanced photoacoustic gas analyser response time and impact on accuracy at fast ventilation rates during multiple breath washout.

Horsley A, Macleod K, Gupta R, Goddard N, Bell N - PLoS ONE (2014)

Bottom Line: A series of previously reported and novel enhancements were made to the gas analyser to produce a clinically practical system with a reduced response time.Signal alignment is a critical factor.With these enhancements, the Innocor analyser exceeds key technical component recommendations for MBW apparatus.

View Article: PubMed Central - PubMed

Affiliation: Institute of Inflammation and Repair, University of Manchester, Manchester, United Kingdom; Manchester Adult Cystic Fibrosis Centre, University Hospital of South Manchester, Manchester, United Kingdom.

ABSTRACT

Background: The Innocor device contains a highly sensitive photoacoustic gas analyser that has been used to perform multiple breath washout (MBW) measurements using very low concentrations of the tracer gas SF6. Use in smaller subjects has been restricted by the requirement for a gas analyser response time of <100 ms, in order to ensure accurate estimation of lung volumes at rapid ventilation rates.

Methods: A series of previously reported and novel enhancements were made to the gas analyser to produce a clinically practical system with a reduced response time. An enhanced lung model system, capable of delivering highly accurate ventilation rates and volumes, was used to assess in vitro accuracy of functional residual capacity (FRC) volume calculation and the effects of flow and gas signal alignment on this.

Results: 10-90% rise time was reduced from 154 to 88 ms. In an adult/child lung model, accuracy of volume calculation was -0.9 to 2.9% for all measurements, including those with ventilation rate of 30/min and FRC of 0.5 L; for the un-enhanced system, accuracy deteriorated at higher ventilation rates and smaller FRC. In a separate smaller lung model (ventilation rate 60/min, FRC 250 ml, tidal volume 100 ml), mean accuracy of FRC measurement for the enhanced system was minus 0.95% (range -3.8 to 2.0%). Error sensitivity to flow and gas signal alignment was increased by ventilation rate, smaller FRC and slower analyser response time.

Conclusion: The Innocor analyser can be enhanced to reliably generate highly accurate FRC measurements down at volumes as low as those simulating infant lung settings. Signal alignment is a critical factor. With these enhancements, the Innocor analyser exceeds key technical component recommendations for MBW apparatus.

Show MeSH

Related in: MedlinePlus

Diagram of lung model.A high-precision computer-controlled linear motor was used to drive a calibration syringe. This moved air into and out of the ventilation compartment of the lung tank. Water level in the lung compartment determined functional residual capacity (FRC) of the model whilst ventilation rate and volume were controlled by speed and excursion of the linear motor. A flowmeter and gas sample needle were connected to the lung compartment. During washin, a T-piece connected the flowmeter to the open circuit 0.2% SF6.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4043749&req=5

pone-0098487-g002: Diagram of lung model.A high-precision computer-controlled linear motor was used to drive a calibration syringe. This moved air into and out of the ventilation compartment of the lung tank. Water level in the lung compartment determined functional residual capacity (FRC) of the model whilst ventilation rate and volume were controlled by speed and excursion of the linear motor. A flowmeter and gas sample needle were connected to the lung compartment. During washin, a T-piece connected the flowmeter to the open circuit 0.2% SF6.

Mentions: The lung model was based on one originally described by Brunner et al. and adapted for use in MBW device validation by Singer et al.[8], [23]. The model consisted of a sealed clear acrylic tank divided into two equal and communicating compartments, as illustrated in Figure 2. The internal dimensions of each compartment were 100×100×125 mm. The tank was filled with water to a pre-set level, according to the desired FRC. The “lung” compartment of the model was connected, via a bacterial filter, to the patient interface of the Innocor device. Pre-capillary deadspace (connector, filter, flowmeter) was calculated at 59 ml (combination of water fill volume and manufacturer-stated volumes, adjusted for connector overlap). The ventilation compartment of the model was connected via a filter to a 1 L calibration syringe (Hans Rudolph Inc, Kansas, USA) attached to a linear actuator (Bosch Rexroth AG, Schweinfurt, Germany). The position, speed and acceleration of the linear motor were controlled by a laptop computer running the manufacturer's own software. Repeatability of the linear motor is 0.2 mm (manufacturer's own data), which represents 0.06 ml of the 1 L syringe volume. Further data on lung model precision are given in the supporting information file S1. A distinct advantage of this over the previously described lung model, which drove the ventilator compartment by changing the pressure settings on a clinical ventilator [8], [20], is that this allows very precise adjustment of the speed and waveform of the breathing pattern as well as precise identification and accurate replication of the end expiratory water level. Accuracy of water level identification was estimated to be within 0.5 mm, reading from a ruler affixed to the internal wall of the lung compartment, or 6.25 ml. In this in vitro system, the model was not heated. Temperature and humidity of both expired air and that within the lung chamber were compared using a digital thermohygrometer (ATP Instrumentation Ltd, Leicestershire, UK), and were found to be the same, so no additional volume correction was applied.


Enhanced photoacoustic gas analyser response time and impact on accuracy at fast ventilation rates during multiple breath washout.

Horsley A, Macleod K, Gupta R, Goddard N, Bell N - PLoS ONE (2014)

Diagram of lung model.A high-precision computer-controlled linear motor was used to drive a calibration syringe. This moved air into and out of the ventilation compartment of the lung tank. Water level in the lung compartment determined functional residual capacity (FRC) of the model whilst ventilation rate and volume were controlled by speed and excursion of the linear motor. A flowmeter and gas sample needle were connected to the lung compartment. During washin, a T-piece connected the flowmeter to the open circuit 0.2% SF6.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4043749&req=5

pone-0098487-g002: Diagram of lung model.A high-precision computer-controlled linear motor was used to drive a calibration syringe. This moved air into and out of the ventilation compartment of the lung tank. Water level in the lung compartment determined functional residual capacity (FRC) of the model whilst ventilation rate and volume were controlled by speed and excursion of the linear motor. A flowmeter and gas sample needle were connected to the lung compartment. During washin, a T-piece connected the flowmeter to the open circuit 0.2% SF6.
Mentions: The lung model was based on one originally described by Brunner et al. and adapted for use in MBW device validation by Singer et al.[8], [23]. The model consisted of a sealed clear acrylic tank divided into two equal and communicating compartments, as illustrated in Figure 2. The internal dimensions of each compartment were 100×100×125 mm. The tank was filled with water to a pre-set level, according to the desired FRC. The “lung” compartment of the model was connected, via a bacterial filter, to the patient interface of the Innocor device. Pre-capillary deadspace (connector, filter, flowmeter) was calculated at 59 ml (combination of water fill volume and manufacturer-stated volumes, adjusted for connector overlap). The ventilation compartment of the model was connected via a filter to a 1 L calibration syringe (Hans Rudolph Inc, Kansas, USA) attached to a linear actuator (Bosch Rexroth AG, Schweinfurt, Germany). The position, speed and acceleration of the linear motor were controlled by a laptop computer running the manufacturer's own software. Repeatability of the linear motor is 0.2 mm (manufacturer's own data), which represents 0.06 ml of the 1 L syringe volume. Further data on lung model precision are given in the supporting information file S1. A distinct advantage of this over the previously described lung model, which drove the ventilator compartment by changing the pressure settings on a clinical ventilator [8], [20], is that this allows very precise adjustment of the speed and waveform of the breathing pattern as well as precise identification and accurate replication of the end expiratory water level. Accuracy of water level identification was estimated to be within 0.5 mm, reading from a ruler affixed to the internal wall of the lung compartment, or 6.25 ml. In this in vitro system, the model was not heated. Temperature and humidity of both expired air and that within the lung chamber were compared using a digital thermohygrometer (ATP Instrumentation Ltd, Leicestershire, UK), and were found to be the same, so no additional volume correction was applied.

Bottom Line: A series of previously reported and novel enhancements were made to the gas analyser to produce a clinically practical system with a reduced response time.Signal alignment is a critical factor.With these enhancements, the Innocor analyser exceeds key technical component recommendations for MBW apparatus.

View Article: PubMed Central - PubMed

Affiliation: Institute of Inflammation and Repair, University of Manchester, Manchester, United Kingdom; Manchester Adult Cystic Fibrosis Centre, University Hospital of South Manchester, Manchester, United Kingdom.

ABSTRACT

Background: The Innocor device contains a highly sensitive photoacoustic gas analyser that has been used to perform multiple breath washout (MBW) measurements using very low concentrations of the tracer gas SF6. Use in smaller subjects has been restricted by the requirement for a gas analyser response time of <100 ms, in order to ensure accurate estimation of lung volumes at rapid ventilation rates.

Methods: A series of previously reported and novel enhancements were made to the gas analyser to produce a clinically practical system with a reduced response time. An enhanced lung model system, capable of delivering highly accurate ventilation rates and volumes, was used to assess in vitro accuracy of functional residual capacity (FRC) volume calculation and the effects of flow and gas signal alignment on this.

Results: 10-90% rise time was reduced from 154 to 88 ms. In an adult/child lung model, accuracy of volume calculation was -0.9 to 2.9% for all measurements, including those with ventilation rate of 30/min and FRC of 0.5 L; for the un-enhanced system, accuracy deteriorated at higher ventilation rates and smaller FRC. In a separate smaller lung model (ventilation rate 60/min, FRC 250 ml, tidal volume 100 ml), mean accuracy of FRC measurement for the enhanced system was minus 0.95% (range -3.8 to 2.0%). Error sensitivity to flow and gas signal alignment was increased by ventilation rate, smaller FRC and slower analyser response time.

Conclusion: The Innocor analyser can be enhanced to reliably generate highly accurate FRC measurements down at volumes as low as those simulating infant lung settings. Signal alignment is a critical factor. With these enhancements, the Innocor analyser exceeds key technical component recommendations for MBW apparatus.

Show MeSH
Related in: MedlinePlus