Development of a bleed detection algorithm for a novel portable ultrasound device to aid vascular trauma victims
Abstract/Contents
- Abstract
- Trauma patients require rapid diagnosis and treatment of hemorrhage. In the hospital, experienced sonographers can reliably diagnose vascular injury using duplex ultrasound. In trauma settings, such as the battlefield, a portable ultrasound device with an automated algorithm to detect bleeding would be useful for medics. Thus, an increasing interest in computer-aided bleed detection led to efforts in quantifying sonographic signatures of abnormalities at the site of vascular injury. However, since trauma patients often present with large areas of injury, there is a need to develop and evaluate bleed detection strategies for ultrasound that more efficiently assess a large vascular tree. The goal of this PhD research dissertation is to address this need. The studies primarily focused on the upper extremity vasculature, specifically the brachial bifurcation, because the leading cause of preventable deaths due to vascular trauma has been exsanguinations from extremity injuries. The overall approach was to characterize normal blood flow with a well-established power law model and identify flows that deviate from the model. The power law states that blood flow is proportional to the vessel diameter raised to a power index k, where k is defined by the bifurcation geometry. A bleed detection metric, called the "flow split deviation" (FSD), was defined to quantify the flow deviations from the power law. Validation of this approach was undertaken in four steps. The first involved demonstrating that the power law model appropriately describes the normal brachial bifurcation and flows in man. The utility of the bleed metric was then evaluated with 3D computational models of bleeds. Finally, the proposed detection algorithm was applied on the early proof of concept humans in arteriovenous fistulas (AVF) of dialysis patients and in in vivo rabbit bleed models. A study with normal human subjects was used to determine that the best-fit k for the brachial bifurcation was 2.75, which is in agreement with other vasculatures previously studied. A k=2.75 power law was then shown to adequately predict forearm blood flows for both resting and exercise physiological states. The correlation coefficient R between predicted and measured normal flows was 0.98. Computational models suggested that FSD was a good indicator of the severity of bleed downstream from the bifurcation. In the patient case study, the bleed metric easily distinguished between normal arms and those with newly placed wrist AVFs, which caused on average an order of magnitude increase in flow deviations. Introduction of different femoral bleed rates in rabbits demonstrated good sensitivity and specificity of the bleed metric when applied to moderate lower extremity bleeds. Bifurcation FSDs can serve as a quantitative signature of bleeding and, moreover, as a strategic way to survey large vascular trees by following abnormal branch points to the likely source of hemorrhage. This approach can complement other quantitative sonographic methods to create a comprehensive, automated, ultrasound-based algorithm for vascular trauma detection.
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 | Wang, Aaron Sheng-Chieh |
|---|---|
| Associated with | Stanford University, Department of Bioengineering. |
| Primary advisor | Taylor, Charles A. (Charles Anthony) |
| Thesis advisor | Taylor, Charles A. (Charles Anthony) |
| Thesis advisor | Liang, David Henchi |
| Thesis advisor | Zarins, Christopher K |
| Advisor | Liang, David Henchi |
| Advisor | Zarins, Christopher K |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Aaron Sheng-Chieh Wang. |
|---|---|
| Note | Submitted to the Department of Bioengineering. |
| Thesis | Thesis (Ph. D.)--Stanford University, 2010. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2010 by Aaron Sheng-Chieh Wang
- 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...