Segmented Wavelet Decomposition for Capnogram Feature Extraction in Asthma Classification

Abstract

A feature extraction method from capnograms used for classifying asthma is proposed based on wavelet decomposition. Its computational cost is low and its performance is adequate for classifying asthma in real time. Experiments performed using 23 capnograms from an asthma camp in Cuba showed 97.39% best classification accuracy. The time required for a physiological multiparameter monitor to determine the suitable features of capnograms averaged 8 seconds. The proposal is to be used as part of a decision support system for asthma classification being developed by TITECH and TMDU research groups.

On Analyzing Capnogram as a Novel Method for Screening COVID-19: A Review on Assessment Methods for COVID-19

“…These features have been used to quantify differences between the shape of a capnogram in a normal subject and that of a patient with an obstructive or restrictive disease Table 1. Capnogram features related to asthma have been widely explored, and have been correlated with spirometric indexes for discriminating asthmatic from non-asthmatic subjects and estimating asthma severity [86]. While some changes in the morphology of the CO2 waveform can be seen with the naked eye, such as the "shark fin" shape for asthmatic patients, more subtle variations require computation and pattern recognition methods.…”

“…The G-H segment presented the best results, with a sensitivity of 55.71%, specificity of 99.38%, correct error of 86.0%, and error rate of 13.91%. The results show that the terminal indices of the capnogram are highly sensitive to airway obstruction [86]. et al [86] evaluated degrees of asthma severity using capnogram features obtained via the decomposition of a breath cycle into small segments (A-B, C-D, E-F and G-H), and further intermediate parts between the segments (B-C, D-E, F-G and H-A) (Figure 8).…”

“…The results show that the terminal indices of the capnogram are highly sensitive to airway obstruction [86]. et al [86] evaluated degrees of asthma severity using capnogram features obtained via the decomposition of a breath cycle into small segments (A-B, C-D, E-F and G-H), and further intermediate parts between the segments (B-C, D-E, F-G and H-A) (Figure 8). A support vector machine was used to classify asthma severity into six classes using feature vectors.…”

“…The G-H segment presented the best results, with a sensitivity of 55.71%, specificity of 99.38%, correct error of 86.0%, and error rate of 13.91%. The results show that the terminal indices of the capnogram are highly sensitive to airway obstruction [86]. Doğan, Nurettin Özgür, et al [89] analyzed the variation in EtCO2 levels in COPD exacerbations.…”

“…These features may increase or decrease during an asthma attack, thus providing useful information related to asthmatic changes [91]. Modifications in the shape of the capnogram waveform have been analyzed by many researchers, and various features have been proposed as indicators of respiratory disorders, with asthma most common among these [84,86,91]. Moreover, Mieloszyk et al [83] quantitatively analyzed capnogram waveforms to discriminate between COPD and congestive heart failure (CHF), and between COPD and normal patients.…”