Classification of hemodynamics from perfusion MR brain images using noiseless independent factor analysis

  • Authors:

  • Yen-Chun Chou;Michael Mu Huo Teng;Wan-Yuo Guo;Jen-Chuen Hsieh;Yu-Te Wu

  • Affiliations:

  • Institute of Radiological Science, National Yang-Ming University and Integrated Brain Res. Lab., Dept. of Med. Res. and Ed., Taipei Veterans General Hospital, Taipei, Taiwan, ROC;Department of Radiology, Taipei Veterans General Hospital, Taipei, Taiwan, ROC and Faculty of Medicine, National Yang-Ming University, Taipei, Taiwan, ROC;Department of Radiology, Taipei Veterans General Hospital, Taipei, Taiwan, ROC and Faculty of Medicine, National Yang-Ming University, Taipei, Taiwan, ROC;Faculty of Medicine, National Yang-Ming University and Integrated Brain Res. Lab., Dept. of Med. Res. and Ed., Taipei Veterans General Hospital and Inst. of Brain Sci., Natl. Yang-Ming Univ., Taip ...;Institute of Radiological Science, National Yang-Ming University and Integrated Brain Res. Lab., Dept. of Med. Res. and Ed., Taipei Veterans General Hospital, and Inst. of Brain Sci., National Yan ...

  • Venue:
  • SPPRA'06 Proceedings of the 24th IASTED international conference on Signal processing, pattern recognition, and applications
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Dynamic-susceptibility-contrast (DSC) magnetic resonance imaging records signal changes on images when the injected contrast-agent particles pass through a human brain. The temporal signal changes on different brain tissues manifest distinct blood supply patterns which are vital for the profound analysis of cerebral hemodynamics. Under the assumption of the spatial independence among these patterns, noiseless independent factor analysis (NIFA) was first applied to decompose the DSC-MR data into different independent-factor images with corresponding signal-time curves. A major tissue type, such as artery, gray matter, white matter, vein, sinus, and choroid plexus, etc., on each independent-factor image was further segmented out by an optimal threshold. Based on the averaged signal-time curve on the arterial area, the cerebral hemodynamic parameters, such as relative cerebral blood volume (rCBV), relative cerebral blood flow (rCBF), and mean transit time (MTT), were computed and their averaged ratios between gray matter and white matter for normal subjects were in good agreement with those in the literature. Data of a stenosis patient before and after treatment was analyzed and the result illustrates that this method is effective in extracting spatio-temporal blood supply patterns which improves differentiation of pathological and physiological hemodynamics.