Early Diagnosis of Alzheimer's Type Dementia Using Continuous Speech Recognition

One of the most important social problems that many of the developed countries face is the constant rise of the percentage of the elderly population values. A major health issue affecting this part of the population is the appearance of dementia of the Alzheimers type (AD) which is the most common case of dementia, affecting around 50 million people worldwide. In addition, the frequency of the AD related cases is expected to grow three times over the next 50 years. The proposed AD status monitoring system (ADSMS) is processing a person’s speech habits to train itself and extracts specific statistic parameters. After that necessary training process, it constantly monitors and analyzes new spontaneous speech data, in order to classify them. In this way, it is possible for the ADSMS to predict possible signs of AD that need further investigation with the common methods of diagnosis by a physician.