Regresar

Identification of the semantic disconnection in alzheimer’s patients conducted by bayesian algorithms

Abstract:

In recent years efforts to find mechanisms that allow early identification of neurodegenerative disease with an impact on Alzheimer’s cognitive abilities or progress have been a concern of the scientific community and caregivers. For this, we start from the hypothesis, supported by the bibliography of the subject, which states that people with early Alzheimer’s present semantic disconnections between the emotions that is showed in the face and feeling, they are shown by an oral or textual phrase. The key point here is that the caregivers can’t be awaiting all the time to find the number of disconnections, but these can be recorded in video and audio as well as be analyzed automatically. Our proposal is to develop a methodology that is based on a software that detects emotions in the face of the participants developed in our study group and in some Bayesian rhythms that allow to classify the sentimental polarity of the conversational phrases. This methodology allows the comparison of results and obtain the moments of semantic disconnection when there is no coincidence between the emotions and the polarity. The experimental results show that it has been possible to identify the disconnections with an 82% success. Our study is an initial proposal, although following previous work that qualifies this line of work.