Research & development - Leuven | More than two weeks ago
Psychomotor agitation and wandering are typical behavioral symptoms seen in patients with dementia. These symptoms can be indicators of patient distress or agitation events and are correlated with disease severity and progression. Modeling these symptoms has the potential to allow just-in-time interventions, eventually improving a patients’ quality of life and decreasing the burden on caregivers. Within a multimodal sensing study on agitation in dementia, a BLE localization system will be used to track a patient’s movements within an institutional ward. The data from this BLE system can be combined with activity data generated by a wrist-worn wearable to investigate the patterns related to wandering and psychomotor agitation in dementia. To enable the characterization of these symptoms, a data pipeline is required to ingest streaming input data from an API, clean the data, engineer features, and apply modeling techniques.
This master thesis aims to develop a tool that will take streamed data as an input and use this tool to model wandering/psychomotor agitation from location and movement-based data. The student will be involved in setting up an end-to-end data pipeline enabling raw location data input to be engineered into relevant features ready for input into machine learning models, as well as the feature engineering and modelling itself.
Specific thesis objectives:
Type of project: Thesis
Duration: 6-9 months
Required degree: Master of Engineering Technology, Master of Science, Master of Engineering Science, Master of Bioengineering
Required background: Biomedical engineering, Bioscience Engineering, Computer Science, Electrotechnics/Electrical Engineering
Supervising scientist(s): For further information or for application, please contact: Hannah Davidoff (Hannah.Davidoff@imec.be)
Imec allowance will be provided for students studying at a non-Belgian university.