The effect of software post-processing applications on identification of the penumbra and core within the ischaemic region in perfusion computed tomography

Purpose
Assessment of software applications designed for post-processing of CT imaging data and perfusion maps in terms of their ability to consistently define the penumbra and core in an ischemic area.

Material and methods
This study is based on measurements conducted in a group of 65 patients with neurological symptoms suggestive of ischaemia in the area of the MCA within 12 hours following onset of the first symptoms. Non-contrast and perfusion CT were performed during an emergency duty. The acquired data were processed using various programs to obtain defined ischaemic areas and parameters. Finally, the results obtained were compared to the distribution of penumbra and core within the ischaemic area received from different perfusion mapping programs.

Results
The programs designed to convert the acquired data and to map the distribution of perfusion were also assessed for their viability in dividing the ischaemic zone into penumbra and core. There was a statistically strong correlation (0.784-0.846) between results obtained by processing of imaging data with two different packages, and then by post-processing with one package, and a poor correlation (0.315-0.554) between results obtained by processing of data with the same package, and post-processing with two different packages designed for measuring penumbra and core areas.

Conclusions
The results obtained by processing of imaging data with different software applications and by post-processing with one program developed for identifying penumbra and core areas show a strong correlation. However, the results obtained by processing imaging data with the same software application and by post-processing with different programs measuring penumbra and core areas reveal poor correlation.