ABSTRACT
Discourse investigation could give a marker of Alzheimer’s dementia sickness and assist with creating clinical devices for naturally distinguishing and checking illness movement. While past examinations have utilized acoustic (discourse) features for characterisation of Alzheimer’s dementia, these investigations zeroed in on a couple of normal prosodic components, regularly in blend with lexical and syntactic elements which require record. A flow learns investigated the utilization of (CA-Conversation Analysis) of meetings among patients and nervous system specialists as an approach to recognize among patients with progress neurodegenerative memory infection (ND) and those with (Non-Progressive) FMD (Functional Memory Disease) to improve dementia acknowledgment accuracy. Manual CA, on the opposite side, is expensive and complex to increase for successive clinical use. In this article, we propose an early dementia detection system utilizing discourse acknowledgment and investigation dependent on NLP method utilizing bidirectional Long Short-Term Memory (bi-LSTM) structure neural architecture design which surprisingly catches the worldly provisions and long- term dependencies from verifiable information to demonstrate the abilities of arrangement models over a feed-forward neural architecture in estimating discourse examination related issues.
