An image analysis on MR imaging of the brain for hepatic encephalopathy

Biomed Sci Instrum. 1991:27:29-33.

Abstract

Through the Magnetic Resonance Imaging (MRI) sequences with T1 weighting (TR 600 msec/TE 20 msec) of the brain, the high intensity lesions due to hepatic encephalopathy have been demonstrated markedly. The data have been collected were based upon eleven patients with clinically proven hepatic encephalopathy. Three sets of normal brain MRI which were identified with no evidence of brain or liver diseases were also included in this study. After the brain MRI slices have been identified, it showed that the affected regions of interest (ROI) were around the largest third ventricle. The histogram which was generated out of that particular slice is composed of two distinguished peaks lumped together. After the decomposition and curve fitting procedures, the Gaussian distributions were identified with the mean value and the standard deviation. One was assumed to be gray index region for gray and white matter, the other with higher gray index value was the contribution of hepatic encephalopathy. Once the range of the distribution has been determined, the affected portion can be marked and highlighted. This mechanism of histogram decomposition and curve fitting techniques shows the valuable correlation with the clinical status of MR findings.

MeSH terms

  • Brain / pathology*
  • Hepatic Encephalopathy / diagnosis*
  • Humans
  • Image Processing, Computer-Assisted*
  • Magnetic Resonance Imaging*
  • Microcomputers