Hemodynamic segmentation of MR brain perfusion images using independent component analysis, thresholding, and Bayesian estimation

Abstract

Dynamic-susceptibility-contrast MR perfusion imaging is a widely used imaging tool for in vivo study of cerebral blood perfusion. However, visualization of different hemodynamic compartments is less investigated. In this work, independent component analysis, thresholding, and Bayesian estimation were used to concurrently segment different tissues, i.e., artery, gray matter, white matter, vein and sinus, choroid plexus, and cerebral spinal fluid, with corresponding signal-time curves on perfusion images of five normal volunteers. Based on the spatiotemporal hemodynamics, sequential passages and microcirculation of contrast-agent particles in these tissues were decomposed and analyzed. Late and multiphasic perfusion, indicating the presence of contrast agents, was observed in the choroid plexus and the cerebral spinal fluid. An arterial input function was modeled using the concentration-time curve of the arterial area on the same slice, rather than remote slices, for the deconvolution calculation of relative cerebral blood flow. Magn Reson Med 49:885-894, 2003.

Blind Source Separation of Concurrent Disease-Related Patterns from EEG in Creutzfeldt–Jakob Disease for Assisting Early Diagnosis

“…ICA has been successfully applied to remove non-physiological artifacts from EEG data, 14,15 to segregate Rolandic beta rhythm from magnetoencephalographic (MEG) measurements of the right index finger lifting, 17 to extract the taskrelated features from the motor imagery EEG and the flash visual evoked EEG in the studies of the brain computer interface, 10,18 to analyze the interactions during temporal lobe seizures in stereotactic depth EEG, 22 to separate generalized spike-and-wave discharges into the primary and secondary bilateral synchrony, 13 and to segment spatiotemporal hemodynamics from perfusion magnetic resonance brain images. 16 …”

Hemodynamic segmentation of MR perfusion images in patients with unilateral carotid stenosis using independent component analysis

“…Fourth, the spatial distributions of output IC images are statistically independent. The theoretical derivation can be found in the literature (28,29).…”

“…A summation of all 65 temporal points on the concentration-time curve was used to calculate the rCBV value (31). The rCBF, MTT, and T max values were calculated using the deconvolution technique proposed by Ostergaard et al (14,15) and an in-plane AIF selected on the normal side (29). The differences in TTP, T max , and MTT of arteries and brain parenchyma on the normal and stenotic sides were compared with the blood supply to MCA territory observed on the cerebral angiograms.…”

“…It is used to separate mixed signals into statistically independent source signals. In applications to dynamic MR images, IC analysis provides an estimate of the partial volume mixing problem (28,29). The segmentation of MR perfusion images of normal subjects using IC analysis provides spatiotemporal hemodynamics of normal brain tissues and an in-plane AIF can be measured for the deconvolution calculation of rCBF (29).…”

The Brain Computer Interface Using Flash Visual Evoked Potential and Independent Component Analysis

“…In the second study, randomly-generated Gaussian noise were added to the ICA-extract FVEP. Eleven levels of signal-to-noise ratio (SNR), ranging from 0, 5,5,5,5,7,8,9,13,15,22, and 25 trials, respectively, to achieve 90% accuracies, which correspond to 2.33, 2.33, 2.33, 2.33, 3.23, 3.72, 4.20, 6.06, 7.00, 10.26, and 11.66 s command transfer interval (CTI), respectively. The same noise levels were also tested in the FMFVEP system ( Fig.…”