This paper is a groundwork research that aims to enhance the accuracy of low-cost ambulatory diagnosis of dementia by augmenting Verbal Fluency Test (VFT) with single-channel Electroencephalography (EEG) measurement. We developed a methodology to interpret EEG signals at Fp1 of the frontal lobe during VFT in order to detect whether the EEG signal suggests dementia. We hypothesized that the difference in EEG patterns between (1) normal controls and (2) dementia patients with damaged semantic networks would show similar characteristics to the difference in EEG patterns between (1) normal controls performing “easy” VFT tasks and (2) normal controls performing “hard” VFT tasks. We then identified EEG indexes that represent the differences of VFT difficulty. A Support Vector Machine (SVM) was used to verify the applicability of the identified EEG indexes. In our experiments, the pairs of EEG samples from 35 subjects were collected from both easy and hard VFT tasks, and our results showed a 65% diagnostic accuracy which is significantly higher than chance. Our research proposes that single-channel EEG measurement can complement VFT to enhance the diagnostic accuracy of dementia.