Skip to main navigation Skip to search Skip to main content

Voxel-based correlation between whole-brain CT perfusion with 320-row area detector CT and iodine 123 iodoamphetamine brain perfusion SPECT in patients with cerebrovascular disease

Research output: Contribution to journalArticlepeer-review

Abstract

Objective: We compared cerebral blood flow (CBF) measured using computed tomographic (CT) perfusion (CTP) and N-isopropyl-p-[(123) I]-iodoamphetamine cerebral perfusion single-photon emission computed tomography (SPECT).

Methods: We used a 320-row area detector CTand N-isopropyl-p-[(123) I]-iodoamphetamine cerebral perfusion SPECTunder similar conditions in patients with chronic cerebrovascular disease. Images were automatically aligned 3-dimensionally for voxel-by-voxel comparisons.

Results: Linear positive correlations were observed between CTPCBF including high-blood-flow areas and SPECT-CBF over the whole brain (r = 0.001-0.6, P < 0.01), superior cerebral level (r = 0.45-0.93, P < 0.01), basal ganglia level (r = 0.44-0.77, P < 0.01), and skull base (r = 0.02-0.66, P < 0.01). Correlations between CTP-CBF excluding high-blood-flow areas were significantly higher (P < 0.0001).

Conclusions: Computed tomographic perfusion overestimated CBF compared with SPECT and showed poor correlation at the skull base. Computed tomographic perfusion CTP excluding high-blood-flow areas improved the correlation over the whole brain in patients with chronic cerebrovascular disease.

Original languageEnglish
Pages (from-to)639-646
Number of pages8
JournalJournal of Computer Assisted Tomography
Volume38
Issue number5
DOIs
Publication statusPublished - 2014
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Radiology Nuclear Medicine and imaging

Fingerprint

Dive into the research topics of 'Voxel-based correlation between whole-brain CT perfusion with 320-row area detector CT and iodine 123 iodoamphetamine brain perfusion SPECT in patients with cerebrovascular disease'. Together they form a unique fingerprint.

Cite this