- Title Information
- Title
- Biomechanistic and Bioinformatic Modeling for Diabetes
- Type of Resource (primo)
- dissertations
- Name:
Personal
- Name Part
- Deng, Yixiang
- Role
- Role Term:
Text
- creator
- Name:
Personal
- Name Part
- Karniadakis, George
- Role
- Role Term:
Text
- Advisor
- Name:
Personal
- Name Part
- Zenit, Roberto
- Role
- Role Term:
Text
- Reader
- Name:
Personal
- Name Part
- Srivastava, Vikas
- Role
- Role Term:
Text
- Reader
- Name:
Personal
- Name Part
- Mantzoros, Christos
- Role
- Role Term:
Text
- Reader
- Name:
Corporate
- Name Part
- Brown University. Engineering: Fluids and Thermal Sciences
- Role
- Role Term:
Text
- sponsor
- Origin Information
- Copyright Date
- 2021
- Physical Description
- Extent
- , None p.
- digitalOrigin
- born digital
- Note:
thesis
- Thesis (Ph. D.)--Brown University, 2021
- Genre (aat)
- theses
- Abstract
- This dissertation focuses on biomechanistic and bioinformatic modeling of diabetes. Specifically, it is expanded on four fronts:
1) Quantification of fibrinogen-dependent red blood cells adhesion. On this front, I integrate companion microfluidic experiments that provide in vitro quantitative information on cell-cell adhesive dynamics, to quantify the rouleau dissociation dynamics using dissipative particle dynamics.
2) Blood cells dynamics in coagulation cascade in normal and diabetic blood. I develop a computational framework that is able to integrate seamlessly four key components of blood clotting, namely transport of coagulation factors, coagulation kinetics, blood cell mechanics and platelet adhesive dynamics, to model the development of thrombi under physiologic conditions and type 2 diabetes.
3) Short-term glucose prediction and long-term hypoglycemia detection. I develop deep learning methods to predict patient-specific blood glucose during various time horizons in the immediate future using patient-specific every 30-min long glucose measurements by the continuous glucose monitoring (CGM) to predict future glucose levels in 5 minutes to 1 hour. For long-term hypoglycemia detection, I utilize multi-view machine learning algorithms to improve accuracy and sensitivity. In both tasks, we are faced with challenges of small dataset and data imbalance, which is commonly seen in real-world data, especially medical data. I will talk about our strategies to tackle the problems.
4) Artificial pancreas design using offline reinforcement learning. I present a novel framework which combines three key components to automating insulin dosage for patients with type 1 diabetes. These important components are a real-world historical medical dataset, a flexible ODEs model defining glucose-insulin dynamics and a offline reinforcement learning algorithm, which is capable of optimizing insulin dosage without interaction with the target environment.
- Subject (fast)
(authorityURI="http://id.worldcat.org/fast", valueURI="http://id.worldcat.org/fast/00892147")
- Topic
- Diabetes
- Subject (fast)
(authorityURI="http://id.worldcat.org/fast", valueURI="http://id.worldcat.org/fast/00832181")
- Topic
- Bioinformatics
- Subject
- Topic
- Computational modeling
- Subject (fast)
(authorityURI="http://id.worldcat.org/fast", valueURI="http://id.worldcat.org/fast/00955045")
- Topic
- Hemodynamics
- Language
- Language Term (ISO639-2B)
- English
- Record Information
- Record Content Source (marcorg)
- RPB
- Record Creation Date
(encoding="iso8601")
- 20220118