Speech analysis for the early detection of Alzheimer’s

2015/03/30 Elhuyar Zientzia Iturria: Elhuyar aldizkaria

• Teknologia
• Medikuntza
• Softwarea

The UPV ELEKIN research group is developing non-invasive methodologies for early detection of Alzheimer's disease. Biomedical engineering is one of the research lines of the group. In particular, biosignal analysis systems are being developed that help medical diagnoses.

Other projects include the automatic analysis of spontaneous speech (Automatic Spontaneous Speech Analysis, ASSA).Based on this, they aim to develop a non-invasive technique of intelligent diagnosis for early detection of Alzheimer's disease.

In this methodology, patients are recorded in a relaxed and pleasant environment while they tell some of the experiences of their life, measuring some parameters. For example, the interruptions that the patient makes when speaking try to remember the word that means are measured. Therefore, the technique is performed without altering or blocking the patient's abilities, since this test does not generate stress in the patient.

The researchers of the ELEKIN group, in collaboration with other entities that are part of the project, carry out various pilot sessions for the development of the same technique and the measurement and optimization of its usefulness. The tests are done to people who have not developed the disease, to people with a family history, and to patients. Thanks to this collaboration, "people can be studied who are in the preclinical phase of the disease or who have not yet developed it," said Karmele López de Ipiña, head of the research.

The ASSA system is not yet available, it is still developing and researching the technique. "We are going to have a system of evaluation of patients that can be used normally," said Ipiña-. We want to provide doctors with tools to detect elements that are not seen at first sight. The automatic revision of these signals will help them."