Predavanje Monitoring cognitive decline through vocal features

Helena Haque Chowdhury | 17. 9. 2017

Slovensko fiziološko društvo vljudno vabi na predavanje, ki ga bo imel

dr. Saturnino Luz,

v petek, 22. septembra 2017 ob 14:00 uri
v srednji predavalnici Medicinske fakultete,
Korytkova 2, v Ljubljani.

Povzetek
Automatic monitoring of physical and cognitive well-being has become a focus of great research interest as well as practical relevance in the area of elderly care. Emerging applications aimed at promoting healthy ageing and improving care have built on advances in sensor technology and machine learning, which provide opportunities for collection and analysis of vast quantities of personal behavioural data. This talk will discuss methods for the detection of Alzheimer’s type dementia though analysis of vocalisations and turn-taking features. Unlike other approaches, these methods do not rely on transcriptions of the patient’s speech but employ features that can be easily extracted from spontaneous speech. Such methods could potentially open new avenues for continuous mental health monitoring through sensors embedded in everyday technologies and for large scale studies of cognitive decline.

DR. SATURNINO LUZ is a Chancellor’s Fellow in the Usher Institute of Usher Institute of Population Health Sciences and Informatics, Edinburgh Medical School, at the University of Edinburgh. He works in the field of medical informatics, focusing on communication and analysis of time-based modalities in healthcare settings. He investigates inference in high-dimensional data sets, and has applied machine learning and signal processing methods in the analysis of speech and non-verbal interaction patterns for dementia prediction, and other areas such as analysis of multidisciplinary medical team meetings, doctor-patient consultations and tele-medicine. Dr Luz leads the Translational Reseach Group on Technology and Communication at the Edinburgh University’s Centre for Dementia Prevention.

Lep pozdrav,
Helena Chowdhury
SFD


pripeto:


© SiNAPSA 2003-2012