Supervisor: Andrea Facchinetti
Co-supervisor: Dr. Elisa Pellizzari
Co-supervisor Department/Company: Dipartimento di Ingegneria dell’Informazione (DEI)
Creation Date: 09/06/2026 14:55
The project aims to investigate the relationship between psychosocial and behavioral factors, and glycemic control in people with type 1 diabetes through probabilistic graphical models.
In particular, Bayesian network approaches will be employed to integrate multimodal longitudinal data, including ecological momentary assessment (EMA), continuous glucose monitoring (CGM), questionnaire-based psychological assessments, and wearable-derived physiological signals. The study will focus on identifying psychosocial and behavioral patterns associated with impaired glycemic control and increased glucose variability, assessed through CGM-derived metrics such as Time in Range (TIR), hyperglycemic and hypoglycemic exposure, and glucose variability indices.
The overall goal is to improve the understanding of psychosocial determinants of poor glucose control in type 1 diabetes and support the development of more personalized management strategies.
Dataset type: Already acquired data
Dataset description: Dataset description: The dataset was acquired during the DIA-LINK Study – Towards a better understanding of diabetes distress, depression, and poor glycemic control leading to personalized interventions for people with diabetes. The dataset includes approximately 200 individuals with type 1 diabetes monitored across: i) baseline assessment, ii) ambulatory monitoring phase, and iii) follow-up assessment. Data of interest are: • CGM data (4 weeks) • HbA1c • Standardized questionnaires to assess diabetes distress and depression • Ecological Momentary Assessment (EMA) data • Physical activity and heart rate data through Garmin wearable device
List of Methods: Time-series and longitudinal data analysis, statistical analysis of CGM-derived glycemic metrics, Bayesian networks
Elaborazione dei segnali biologici, machine learning for bioengineering