14 March 2013 Stroke prognosis by applying double thresholds on CT-perfusion-brain images
Author Affiliations +
Proceedings Volume 8768, International Conference on Graphic and Image Processing (ICGIP 2012); 876845 (2013) https://doi.org/10.1117/12.2011164
Event: 2012 International Conference on Graphic and Image Processing, 2012, Singapore, Singapore
Abstract
The CT-perfusion image shows information of brain abnormalities such as its size and location. Generally, neurologist diagnoses stroke disease using CT-perfusion images such as Cerebral blood flow (CBF), cerebral blood volume (CBV). In our previous report, we applied threshold technique to divide amount of CBV and CBF into low and high level. Then, their levels are applied to identify normal tissue areas, dead tissue areas (infract core) and blood-cot tissue areas (infract penumbra). However, it’s not totally correct, if the same threshold is applied to the whole area (it must depend on size of blood vessel in that area. In this report, we propose double thresholds to divided CBV and CBF into 3 levels: very low, medium and very high levels. Very low and very high levels are definitely implied to bad areas and good areas, respectively. The proposed double thresholds makes stroke prognosis more accurate. The simulation results confirm that our proposed results closed to results defined from neurologist comparing to the conventional results.
© (2013) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Somchart Chokchaitam, Somchart Chokchaitam, Nittaya Santipromwong, Nittaya Santipromwong, Sombat Muengtaweepongsa, Sombat Muengtaweepongsa, } "Stroke prognosis by applying double thresholds on CT-perfusion-brain images", Proc. SPIE 8768, International Conference on Graphic and Image Processing (ICGIP 2012), 876845 (14 March 2013); doi: 10.1117/12.2011164; https://doi.org/10.1117/12.2011164
PROCEEDINGS
5 PAGES


SHARE
Back to Top