Personalized models of pathophysiology to improve outcome
Abstract/Contents
- Abstract
- Personalized medicine has arisen from the long-recognized fact that drug efficacy and safety vary from person to person. Genetics and genomics have been the main tools of "traditional" personalized medicine, yet the daily practice of physicians revolves around physiology. In this dissertation, I approached personalized medicine using clinical-level physiology and demonstrated this approach in two ways. First I showed how individualized understanding of pathophysiology could improve treatment of a patient with dystonia. Then I demonstrated using multivariate clinical physiology in the intensive care unit (ICU) to develop an understanding of the factors associated with poor outcome and the effects of drugs. Childhood dystonia is a movement disorder often caused by cerebral palsy. Its severity varies from barely noticeable to deeply profound, affecting every aspect of daily life. One application of personalized physiology modeling included the creation of an algorithm to estimate the user's intent when using a computer interface. Aside from doubling the patient's typing rate compared to his normal interface, we were also able to discern individual patterns reflective of his specific pathophysiological movements. Similarly, studying his reaching movement patterns enabled us to test a biofeedback protocol to train him to make better movements, resulting in straighter movements more resembling normal reaching. Thus, leveraging knowledge of individual patients led to speedy improvements over conventional treatments and improved outcome. Having demonstrated the applicability of these methods to two specific problems, I then approached the more general problem of modeling critical care patients in the intensive care unit. Physiology monitoring in the ICU has been common, but simplistic, in most cases using only single variables to define alarms to prevent catastrophic consequences. Using temporally dense multivariate data from the ICU, I discussed the definition of complex physiological states and associated them with the likelihood of poor outcomes. Furthermore, examining pairwise relationships between variables in the complex states most and least associated with survival suggested that mitochondrial dysfunction could be involved in the poor outcomes. Further examination of the relationships between variable pairs led to the construction of physiological correlation networks associated with the presence of infection and the administration of drugs. In both of these comparisons, clear differences were evident between the two cases. This demonstrated that physiological relationships depend on patient state and suggest further study to clarify them for implementation in patient monitoring routines. In this dissertation I demonstrated, using both a pilot study of critical care pathophysiology models and a concrete application of personalized modeling to pediatric dystonia patients, that personalized medicine could benefit by incorporating pathophysiology into its modeling efforts. Further developing an understanding of how patients transition between states predictive of good and poor outcomes will enable improvements in critical care treatment and beyond.
Description
| Type of resource | text |
|---|---|
| Form | electronic; electronic resource; remote |
| Extent | 1 online resource. |
| Publication date | 2010 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Associated with | Grossman, Adam David | |
|---|---|---|
| Associated with | Stanford University, Department of Bioengineering. | |
| Primary advisor | Butte, Atul J | |
| Thesis advisor | Butte, Atul J | |
| Thesis advisor | Altman, Russ | |
| Thesis advisor | Sanger, Terence D. (Terence David) | |
| Advisor | Altman, Russ | |
| Advisor | Sanger, Terence D. (Terence David) |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Adam David Grossman. |
|---|---|
| Note | Submitted to the Department of Bioengineering. |
| Thesis | Thesis (Ph. D.)--Stanford University, 2010. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2010 by Adam David Grossman
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...