Detection of cigarette smoke inhalations from respiratory signals using reduced feature set

Patil, Yogendra; Lopez-Meyer, Paulo; Tiffany, Stephen T.; Sazonov, Edward

doi:10.1109/embc.2013.6610927

Cited by 9 publications

(11 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Random forests exhibited higher precision rates on average than both AdaBoost and bagging but at the expense of lower recall ability. Average performance (across all model types) was higher than our previous group results computed using the extracted waveforms but lower than results found using 1503element feature vectors (Table V, [10]). Figure 2 illustrates the performance of the Random Forest group models by plotting several ROC curves.…”

Section: Discussioncontrasting

confidence: 78%

“…2) Explore the utility of feature selection methods (utilized in previous work [10]) for decision tree ensemble models.…”

Section: Discussionmentioning

confidence: 99%

“…The airflow signal is defined as the first derivative of the processed tidal volume V T (t) signal [9] and calculated: [10] 27 descriptive features were extracted from each hand-tomouth gesture and are listed in Table I. These are described in more detail in [10], [8].…”

Section: Signal Pre-processingmentioning

confidence: 99%

“…Models 1) Decision Trees: A decision tree is a predictive model that recursively partitions observations into leaves that are successively more pure (as measured using a variety of methods) [11]. Branches of the tree represent splits of the data by Table I EXTRACTED FEATURES [10] Feature Description 1 duration of expiration 2 time-duration from the start time of a hand gesture to peak of AS(t) 3 time-duration of hand-to-mouth gesture to point of air-flow exceeding threshold 4 time-duration of hand remaining near mouth 5 peak inspiration following the hand-to-mouth gesture (AS(t)) 6…”

Section: Signal Pre-processingmentioning

confidence: 99%

See 3 more Smart Citations

Detection of cigarette smoke inhalations from respiratory signals using decision tree ensembles

2015

Self Cite

View full text Add to dashboard Cite

In this study we explored the ability of ensembles of decision trees to classify hand-to-mouth gestures in order to detect cigarette smoke inhalations. Three subject independent models were constructed using a variety of ensemble techniques: boosting (AdaBoost), bootstrap aggregating (bagging), and Random Forests. Data was gathered during previous studies by extracting features from the signal waveforms of worn sensors. Each hand gesture was associated with either a smoke inhalation or a hand gesture of another type (e.g. eating). Subject as well as group models were trained. For the group models, model performance was evaluated by computing F-score, precision, and recall statistics using a 20-fold leave-one-out cross-validation testing strategy where one subject was held out for evaluation and models were trained on the remaining 19 subjects. For the individual models, models were trained on a single subject and evaluated using 5-fold cross validation. The average and standard deviation of each statistic across all folds were reported.

show abstract

Section: Discussioncontrasting

confidence: 78%

“…2) Explore the utility of feature selection methods (utilized in previous work [10]) for decision tree ensemble models.…”

Section: Discussionmentioning

confidence: 99%

Section: Signal Pre-processingmentioning

confidence: 99%

Section: Signal Pre-processingmentioning

confidence: 99%

See 2 more Smart Citations

Detection of cigarette smoke inhalations from respiratory signals using decision tree ensembles

2015

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the first step, the smoker puffs the cigarette, drawing smoke into the mouth. In the second step, the smoker inhales the smoke (Baker & Dixon, 2006 ; Herning et al, 1983 ; Patil et al, 2013 ; USDHHS, 1988 ).…”

Section: Introductionmentioning

confidence: 99%

Effect of cigarette design on biomarkers of exposure, puffing topography and respiratory parameters

Appleton

Liu

Lipowicz

et al. 2015

Inhalation Toxicology

View full text Add to dashboard Cite

Despite the lack of evidence, many reports exist which have implied that smokers inhale low-yield cigarette smoke more deeply than that of high-yield cigarettes. The objective of this study was to investigate the effect of short-term switching between smoker’s own brand and test cigarettes with different smoke yields on puffing topography, respiratory parameters and biomarkers of exposure. Participants were randomly assigned to smoke either a Test Cigarette-High Tar (TCH), for two days, and then switched to a Test Cigarette-Low Tar (TCL), for two days or the reverse order (n = 10 each sequence). Puffing topography (CReSS microdevice), respiratory parameters (inductive plethysmography) and biomarkers of exposure (BOE, urinary nicotine equivalents – NE and blood carboxyhemoglobin – COHb) were measured at baseline and on days 2 and 4. The average puffs per cigarette, puff volume and puff durations were statistically significantly lower, and inter-puff interval was significantly longer for the TCH compared to the TCL groups. Respiratory parameters were not statistically significantly different between the TCH and TCL groups. Post-baseline NE and COHb were statistically significantly lower in the TCL compared to the TCH groups. Under the conditions of this study, we found no indication of changes in respiratory parameters, particularly inhalation time and volume, between study participants smoking lower versus higher yield cigarettes. Likewise, the BOE provides no indication of deeper inhalation when smoking low- versus high-yield cigarettes. These findings are consistent with the published literature indicating smoking low-yield cigarettes does not increase the depth of inhalation.

show abstract

A New Approach for Smoking Event Detection Using a Variational Autoencoder and Neural Decision Forest

Fan

Gao

2020

IEEE Access

View full text Add to dashboard Cite

Smoking is associated with cancer, cardiovascular disease and premature death and can cause severe fire hazards. To assist with smoking cessation, this paper presents a wireless body area networkbased system consisting of two off-the-shelf devices, one smartphone and one smartwatch, to detect smoking events by mining the inertial sensor data from both devices. In the system, an end-to-end trainable unified model is implemented by combining a variational autoencoder with a random forest to classify the collected data into smoking and nonsmoking after data preprocessing. The variational autoencoder is adopted to learn the feature representation and deal with the class imbalance problem, and the stochastic decision forest is adopted to guide the global optimization of the parameter learning process. Extensive validation of our scheme is performed on a large dataset we collected, and our scheme yields quite promising results in terms of accuracy and efficiency.

show abstract

Detection of cigarette smoke inhalations from respiratory signals using reduced feature set

Cited by 9 publications

References 10 publications

Detection of cigarette smoke inhalations from respiratory signals using decision tree ensembles

Detection of cigarette smoke inhalations from respiratory signals using decision tree ensembles

Effect of cigarette design on biomarkers of exposure, puffing topography and respiratory parameters

A New Approach for Smoking Event Detection Using a Variational Autoencoder and Neural Decision Forest

Contact Info

Product

Resources

About