Neural Network Analysis of the Volumetric Capnogram to Detect Pulmonary Embolism

Patel, Manish; Rayburn, Daniel B.; Browning, Jane A.; Kline, Jeffrey A.

doi:10.1378/chest.116.5.1325

Cited by 19 publications

(16 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Earlier in the 1960s and 1970s, authors focused on the potential diagnostic strength of the difference between P a CO 2 and expired end-tidal CO 2 (EtCO 2 ), expressing alveolar deadspace from well-ventilated but unperfused alveoli, but pitfalls and sources of errors caused this test to be abandoned [10][11][12][13][14]. The interest in CO 2 re-emerged in the late 1990s, when authors studied the combination of D-dimer measurement and volumetric capnography (VCap) for the exclusion of PE [15][16][17][18][19]. VCap, which displays the plot of the expired CO 2 concentration against the expired volume, has theoretical advantages over time-based capnography in measuring deadspace lung volumes and distinguishing obstructive lung diseases from PE [20,21].…”

Section: Introductionmentioning

confidence: 99%

Volumetric or time‐based capnography for excluding pulmonary embolism in outpatients?

Verschuren

Sanchez

Righini

et al. 2010

Journal of Thrombosis and Haemostasis

View full text Add to dashboard Cite

Summary. Background: Volumetric capnography is technically more demanding but theoretically better than the time-based alveolar deadspace fraction (P a CO 2 -EtCO 2 )/P a CO 2 as a bedside diagnostic tool for excluding pulmonary embolism (PE) in outpatients. Objective: We compared both diagnostic accuracy in patients with a suspected PE and positive D-dimer enzymelinked immunosorbent assay results. Patients and methods: In this clinical multicenter trial with prospective inclusion and 3-month follow-up, alveolar deadspace fraction was compared by receiver operating characteristic (ROC) analysis with other parameters derived from volumetric capnography. Results: Capnography was performed in 239 patients, and 205 tests (86%) were conclusive. The incidence of PE was 33%. The alveolar deadspace fraction accuracy expressed with ROC curve analysis was 0.73 ± 0.04. The diagnostic performances of parameters from volumetric capnography were not significantly better. Sixteen per cent [95% confidence interval (CI) 12-21%] of patients presented a (P a CO 2 -EtCO 2 )/P a CO 2 ratio under the cut-off value of 0.15, with a low clinical probability. This combination excluded PE, with a sensitivity of 96% (95% CI 89-99%) and a negative likelihood ratio of 0.17 (95% CI 0.09-0.33%). Conclusion: Volumetric capnography failed to show superiority to alveolar deadspace fraction measurements [(P a CO 2 -EtCO 2 )/P a CO 2 ] for exclusion of PE in outpatients with positive D-dimer test results. Future studies should clarify the safety of excluding PE in patients combining low clinical probability with positive D-dimer results and (P a CO 2 -EtCO 2 )/P a CO 2 ratios below the cut-off value of 0.15.

show abstract

Section: Introductionmentioning

confidence: 99%

Volumetric or time‐based capnography for excluding pulmonary embolism in outpatients?

Verschuren

Sanchez

Righini

et al. 2010

Journal of Thrombosis and Haemostasis

View full text Add to dashboard Cite

show abstract

“…We did not evaluate the slope of capnographic phase 3, which may contain useful information regarding the presence of airway restriction versus blood flow obstruction. Likewise, we did not evaluate variables that require simultaneous measurement of arterial and capnographic partial pressures of carbon dioxide, such as the alveolar dead space volume and the late dead space fraction, that have shown accuracy at predicting pulmonary emboli [21,29]. Paradoxically, this also shows the strength of our GP.…”

Section: Discussionmentioning

confidence: 99%

“…Few studies have used breath-based computer learning methods for predicting pulmonary embolism. Patel et al used a neural network analysis of the volumetric capnogram and found 100% sensitivity and 48% specificity with their model [21]. Neural networks and GP are similar in that they both use datasets to train the computer models to predict outcome.…”

Section: Discussionmentioning

confidence: 99%

Use of genetic programming to diagnose venous thromboembolism in the emergency department

Engoren

Kline

2007

Genet Program Evolvable Mach

Self Cite

View full text Add to dashboard Cite

Pulmonary thromboembolism as a cause of respiratory complaints is frequently undiagnosed and requires expensive imaging modalities to diagnose. The objective of this study was to determine if genetic programming could be used to classify patients as having or not having pulmonary thromboembolism using exhaled ventilatory and gas indices as genetic material. Using a custom-built exhaled oxygen and carbon dioxide analyzer; exhaled flows, volumes, and gas partial pressures were recorded from patients for a series of deep exhalation and 30 s tidal volume breathing. A diagnosis of pulmonary embolism was made by contrast-enhanced computerized tomography angiography of the chest and indirect venography supplemented by 90-day follow-up. Genetic programming developed a series of genomes comprising genes of exhaled CO 2 , O 2 , flow, volume, vital signs, and patient demographics from these data and their predictions were compared to the radiological results. We found that 24 of 178 patients had pulmonary embolism. The best genome consisted of four genes: the minimum flow rate during the third 30 s period of tidal breathing, the average peak exhaled CO 2 during the first 30 s period of tidal breathing, the average peak of the exhaled O 2 during the first 30 s period of tidal breathing, and the average peak exhaled CO 2 during the fourth period of tidal breathing, which immediately followed a deep exhalation. This had 100% sensitivity and 91% specificity on the construction population and 100% and 82%, respectively when tested on the separate validation population. Genetic programming using only data obtained from exhaled breaths was very accurate in classifying patients with suspected pulmonary thromboembolism.

show abstract

“…ANNs have been applied during mechanical ventilation for breathing-pattern recognition during spontaneous breathing and PSV (26); in analysis of pressure and flow waveforms to differentiate normal from injured lungs (27); to determine ventilator settings for neonates (28); to identify respiratory abnormalities by using a pressure monitor to classify breathing patterns as effective or not and to predict changes in arterial oxygen saturation (29); to detect pulmonary embolism (30); and for assessing respiratory system mechanics during ventilatory support (31,32). We contend that another use for computer-based ANNs may be to provide the analytical basis for assessing POB.…”

Section: Discussionmentioning

confidence: 99%

Power of breathing determined noninvasively with use of an artificial neural network in patients with respiratory failure*

Banner

Euliano

Brennan

et al. 2006

Critical Care Medicine

View full text Add to dashboard Cite

POB can be calculated noninvasively with reasonable clinical accuracy for patients receiving ventilatory support by using an ANN. This method obviates the need for inserting an esophageal catheter and thus greatly simplifies measurement of POB. POB(N) may be a clinically useful tool for consideration when setting PSV to unload the respiratory muscles. Before considering its use in clinical practice, POB(N) would need to be incorporated within the context of load tolerance and shown to improve outcomes.

show abstract

Neural Network Analysis of the Volumetric Capnogram to Detect Pulmonary Embolism

Cited by 19 publications

References 37 publications

Volumetric or time‐based capnography for excluding pulmonary embolism in outpatients?

Volumetric or time‐based capnography for excluding pulmonary embolism in outpatients?

Use of genetic programming to diagnose venous thromboembolism in the emergency department

Power of breathing determined noninvasively with use of an artificial neural network in patients with respiratory failure*

Contact Info

Product

Resources

About