Type-1 diabetes is a chronic disease that has a negative impact on the life of a person with diabetes causing other chronic diseases, reducing the quality of life, and the possibility of causing dangerous reductions in blood glucose levels that may lead to coma or death. More than 100 million U.S. adults are now living with diabetes or pre-diabetes. Diabetes is one of the most expensive public health problems in the U.S. at $327 billion in 2017. Thus, alternative solutions or novel proposals are crucial to more effective treatments and cure.
Artificial pancreas systems are one of the common treatment techniques of Type-1 Diabetes, which reduce the risk of diabetes-related complications and make diabetics' lives easier and make it convenient. Artificial pancreas systems aim to maintain blood glucose concentrations in a tighter target blood glucose range, which is a challenging problem.
Several factors affect blood glucose concentrations including intensity of exercise, type of exercise, acute psychological stress and the physical state of a person with diabetes. These factors are unknown disturbances for artificial pancreas control systems. In this project, a single non-invasive wrist-worn device is used to obtain different biosignals in-real time. Biosignals are utilized with the development energy expenditure estimation model, psychological stress detection and physical state classification models. Several machine learning methods are tested and validated until the best classification and estimation accuracy is achieved for each estimate. Obtained models are incorporated with the current artificial pancreas design to improve its glycemic control performance. The controller is aware of such measurable disturbances with the proposed method, which allows for providing accurate and timely control action. Additional estimates are utilized to improve blood glucose concentration prediction model accuracy. Clinical trials are used to test and validate the proposed work.
In summary, the presented work illustrates different machine learning techniques and algorithms that can enhance automated insulin delivery by a multivariable artificial pancreas system and enhance the quality of life of people with Type 1 diabetes.
|Commitee:||Papavasiliou, Georgia, Tichauer, Kenneth, Edirisinghe, Indika|
|School:||Illinois Institute of Technology|
|School Location:||United States -- Illinois|
|Source:||DAI-B 82/1(E), Dissertation Abstracts International|
|Subjects:||Biomedical engineering, Physiology|
|Keywords:||Artificial pancreas, Energy expenditure estimation, Physical state classification, Psychological and physiological state estimation, Psychological stress detection, Wearable device applications|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be