Easy Detection of Rare Dementia Based on Speech Analysis

The increase in the number of dementia patients is a serious problem in developed countries. It is important to diagnose an underlying disease correctly because dementia has treatments depending on the disease. We have been investigating frontotemporal lobar degeneration (FTLD), which is one of the underlying diseases. This paper presents a speech analysis-based FTLD screening system. We used speech features to classify FTLD, Alzheimer's disease (AD) and HC. We confirmed that our method can classify three groups with accuracy of 0.81 and macro F-measure of 0.74. Our screening system has the potential to detect FTLD through short speech.