Stroke-Analysis Program for Searching Infarct/Penumbra Area from CT-Scan Images via Digital Image Processing Technique

摘要

The CT scan, produces x-rays images of brain, can often show size and locations of brain abnormalities. However, with this process, a number of needed CT-scan images prolong the stroke analysis. This report applied digitalimage processing technique to design stroke-analysis program for searching infarct penumbra area (mismatch area), tissue at risk. Cerebral blood flow (CBF), cerebral blood volume (CBV) and Mean Transit Time (MTT) from CT perfusion maps are used to identify mismatch area. In our experimental results, N-Match, D-Match and Mismatch mean normal tissue areas, dead tissue areas and blood cot tissue areas, respectively. The simulation results illustrate percentages of N-Match, D-Match and Mismatch depending on their threshold. For examples, percentage of NMatch will increase when threshold of CBV and CBF are lower and threshold of MTT is higher.