Estudi, desenvolupament i validació d'eines avançades per a la gestió de la diabetis

Compartir
Diabetes mellitus (DM) is a chronic disease characterized by the body's inability to produce the necessary amount of insulin to regulate blood glucose (BG) levels. Most cases occur due to a destruction of the pancreatic beta cells, which are responsible for secreting insulin, the root of an autoimmune process (type 1 DM) or, due to ineffective use of insulin (type 2 DM). Intensive insulin therapy using multiple doses of insulin is the standard of care of T1D patients and allows them to reduce blood glucose levels to prevent long-term diabetes complications. Hypoglycemia is a limiting factor in the application of this therapy and imposes one of the most important challenges in the treatment of people with diabetes, whose aim is to maintain normoglycemia. People with T1D face a lifelong challenge of maintaining BG levels within a safe range by reducing hyperglycemia without provoking hypoglycemic events. There are many mobile health (m-health) applications (apps) for diabetes management and control of BG levels, but most of them do not have tools focused on the most difficult conditions for people with DM, hyperglycemia and hypoglycemia. This is why one of the main motivations of this thesis is not only to improve m-health apps for diabetes self-management decision support, but also to accurately predict hyperglycemia and hypoglycemia events, which would give the patient time to intervene and prevent these BG fluctuations and improve their health and quality of life. Thus, the proposal of the thesis is to design and develop strategies and algorithms for monitoring, analysis and visualization of data, for BG control and improvement of the living conditions of those people living with DM under MDI therapy. The objective of the first part of the study is to develop new tools and functionalities for the assessment of glycemic control and risk of hyperglycemia and hypoglycemia events as a decision support to improve the performance of m-health apps dedicated to DM. The second part of the thesis aims to develop and implement a hypoglycemia prediction and prevention model as a decision support system capable of predicting nocturnal hypoglycemia through the implementation of mitigation strategies using Machine Learning techniques ​
​L'accés als continguts d'aquesta tesi queda condicionat a l'acceptació de les condicions d'ús establertes per la següent llicència Creative Commons: http://creativecommons.org/licenses/by/4.0/

Localització