This paper describes the verification of alcohol detection methodology that measures changes in electrodermal activity (EDA) using electroencephalograms and conductive patch tests. We tested the methodology using a direct current potential method. First, it was revealed that the direct current potential method could detect alcohol in 70% of human subjects by analyzing absolute values of data waves or through a post-processing method. Second, mice that do not sweat were tested to observe the changes in EDA values before and after they were injected with alcohol because EDA might be affected by sweat in humans. Third, to clarify the local mechanism of EDA, fluctuations in ion concentration distribution were analyzed using artificial cells. The results indicate that EDA changed significantly before and after drinking alcohol, with no effects from sweat;​ EDA could be measured before and after alcohol intake;​ and EDA data were available for human skin covered with durable materials. Changes in EDA have been considered to be caused by differences in ion concentrations during transport across the cell membrane, but our experiments using artificial cells revealed that ion concentration was not changed locally by alcohol concentration. It is concluded that drinking alcohol does not cause changes in alcohol concentration by ion transport in the local cell membrane, but has some effect on the transmission system between human cells. However, this requires consideration of alcohol resistance in the human body.

This study clarifies the transmission system between human cells by measuring changes in electroencephalograms before and after drinking alcohol using an electroencephalograph (EEG), and verifies the relationship between EDA and the presence or absence of alcohol resistance by performing a patch test. Changes were observed in the frequency of EEGs before and after drinking alcohol, which suggests that changes in EEGs would have some effect on the transmission system between human cells and cause changes in EDA. In addition, considering the relationship between the patch test and EDA, differences in EDA were small in cases where volunteers were likely to have alcohol resistance, while they were large in cases where volunteers were likely to have no alcohol resistance. Further studies are expected to develop effective driver alcohol detection systems with cost advantages.