Table 2.
Performance of the most notable experimental devices and techniques.
| Author | Device/Technique | Year | Performance Summary | Reference |
|---|---|---|---|---|
| Widmark E.M.P. | Widmark flask | 1918 | First direct measure of ethanol blood concentrations | Early BrAC methods |
| Widmark E.M.P. et al. | EBAC equation | 1924 | Largely inaccurate by modern standards, error in the ranges of ±20% from true value | Widmark Flask and early BrAC methods |
| Brokenstein R.F. et al. | Breathalyzer (photovoltaic assay) |
1961 | Revolutionary device in the field of portable testing devices for intoxication, susceptible to environmental error and variance in lung volume across the population | Analysis of blood and bodily fluids |
| Mishra et al. | THC and ethanol saliva sensing ring | 2020 | Detection range: 0.1 to 1 mM (0.1 mM increments RSD of 1.5% (n = 5) Stable multianalyte sensing (THC) |
Commercial BrAC device |
| Chen et al. | PPG datum line analysis | 2018 | 85% identification rate 18 ms processing and identification time |
Commercial BrAC device |
| Wang et al. | ECG and PPG analysis | 2017 | 95% identification rate Only identifies if a subject is above 0.15 mg/dL |
Commercial BrAC device |
| Rachakonda et al. | Multisensory steering wheel | 2020 | Detection between sober and intoxicated at 0.08 mg/dL Accuracy of 93% |
No reference stated |
| Kubieck et al. | IR facial imaging | 2019 | No specific correlation number states Results indicate a very strong correlation between alcohol consumption and facial temperature distribution in all cases |
No reference stated |
| Chaplik et al. | Bioimpedance spectroscopy | 2019 | Noticeable changes between intoxication and reference group Weak correlation with absolute impedance (r = 0.47) Sensitivity 92% Specificity 76% |
Commercial BrAC device Blood-sample analysis (method unknown) |
| Wen-fei et al. | NIR dynamic spectrum | 2011 | Calibration set: R = 0.9672 Prediction set: R = 0.9384 Relative error between 0.6 and 9%, average error 3.26% |
Hospital biochemical analysis |
| Yamakoshi et al. | Integrated sphere finger-PPG | 2015 | Lower SNR compared to traditional PPG acquisition method Sensitivity of 0.43 ± 0.29 |
No reference (Pilot Study) |
| Kim et al. | Iontophoretic biosensing system | 2016 | Correlation recorded = 0.912 High specificity for ethanol Increased accuracy of the system at higher ethanol concentrations |
FDA-approved commercial BrAC device |
| X. Guo et al. | Wavelength-modulated differential photometry | 2018 | High ethanol resolution: 5–6 mg/dL Lag of 10–15 between ISF and blood ethanol Correlation between 0.96 and 0.98 |
Commercially available BrAC device |
| Lansborp et al. | Wearable enzymatic alcohol biosensor | 2019 | Linear sensor response between 0 and 0.05 mol/L Results of the sensor closely resemble those predicted by Widmark equation, however fall short during the decay stage, and generally underestimate ethanol readings |
Widmark equation (BrAC device deemed impractical for application) |
| Arakawa et al. | Skin ethanol gas | 2020 | Strong correlation of 0.995 Range of estimation 73.9–112.1 ppb/cm2 Results demonstrate superiority over an ordinary biosniffer |
No reference for intoxication measure stated |
| Results indicate strong correlation for at least 3 distinct levels of ethanol | ||||
| Selvam at al. | EtG biochemical sensor | 2016 | Ethanol detection in the range of 0.001–100 ug/L Lower sensitivity at 1 ug/L with gold electrodes compared to ZnO (sensitivity of 0.001ug/L) Three distinct levels of EtG identified Correlation of 0.97 |
|
| Venugopal et al. | ISF sensor for remote continuous alcohol monitoring | 2008 | Generally strong correlation between 0.7203 to 0.866 Correlation between BrAc = 0.879 |
BrAC device and blood testing |
| Tehrani et al. | Microneedle ISF Lactate/Ethanol and Glucose Sensor | 2022 | Low cross-talk between sensing elements Correlation of 0.94 |
Commercially available BrAC device |