Work

Decimation and Analysis: An Internal Bleeding Detection Framework for 3D Ultrasound

Year: 2018
Type: dissertation
Abstract: This thesis proposes a decimation-analysis (D-A) framework for detecting internal bleeding in 3D ultrasound medical scans. The structure consists of two main components: decimation, in which an input volume is reduced to suspect regions, and analysis, in which feature extraction is performed on said suspect regions. The D-A framework was designed to identify intuitive, easily-identiable features of blood pools as an aid for obtaining more complex shape information that can be assessed and matched to "library" data created from training volumes. The D-A framework is presented here for its concept and structure rather than for its actual performance, as rigorous, accurate testing was beyond the scope of the thesis; however, access to accurately segmented ground truth volumes (something missing during the experiment) would enable more in-depth development and validation of the framework.
Language: English
Open Access status: green
OpenAlex ID: W3157544090
Cites: 34
Cited by:
Related to: 10