Reduced hippocampal metabolism in MCI and AD: automated FDG-PET image analysis

L Mosconi1, W-H Tsui, S De Santi

  • 1Department of Psychiatry, New York University School of Medicine, 560 First Avenue, New York, NY 10016, USA.

Neurology
|June 16, 2005
PubMed
Abstract

Insights

A new HipMask technique accurately assesses hippocampal glucose metabolism in mild cognitive impairment (MCI) and Alzheimer disease (AD) using PET scans. This method reveals significant reductions, aiding in diagnostic classification.

Area of Science:

  • Neuroimaging
  • Biomarkers
  • Metabolic Imaging

Background:

  • Automated voxel-based analysis (VBA) in 2-[18F]fluoro-2-deoxy-d-glucose PET (FDG-PET) studies paradoxically fails to detect hippocampal glucose metabolism reductions in mild cognitive impairment (MCI) and Alzheimer disease (AD).
  • Region of interest (ROI) based analyses have reported hippocampal reductions, suggesting limitations in current automated methods for analyzing this specific brain region.
  • The development of an automated, anatomically valid mask technique (HipMask) for hippocampus sampling on PET was necessitated by these discrepancies.

Purpose of the Study:

  • To introduce and validate a novel automated mask technique (HipMask) for accurate hippocampal sampling in PET imaging.
  • To compare the efficacy of the HipMask technique against traditional MRI-guided ROIs and automated VBA for assessing hippocampal glucose metabolism.
  • To investigate the diagnostic utility of hippocampal metabolic changes in distinguishing between healthy elderly (NL), MCI, and AD groups.

Main Methods:

  • The HipMask technique was developed using hippocampal ROIs from MRI scans of 48 subjects (20 NL, 16 MCI, 12 AD).
  • The HipMask was applied to FDG-PET data from 11 NL, 13 MCI, and 12 AD subjects.
  • Comparisons were made between HipMask, MRI-guided ROIs, and VBA methods for analyzing hippocampal and cortical metabolic rates.

Main Results:

  • Both HipMask and ROI methods demonstrated significant and equivalent reductions in hippocampal glucose metabolism when comparing MCI and AD groups to NL.
  • Voxel-based analysis (VBA) identified typical cortical effects but failed to detect hippocampal metabolic reductions in MCI or AD.
  • Hippocampal metabolic rate, analyzed via HipMask, was the sole discriminator between NL and MCI (78% accuracy) and improved classification accuracy for NL vs. AD and MCI vs. AD when combined with cortical data.

Conclusions:

  • The HipMask technique offers a rapid and accurate method for assessing hippocampal glucose metabolism on PET without manual ROI delineation.
  • Significant reductions in hippocampal glucose metabolism are present in both MCI and AD, playing a crucial role in their diagnostic classification.
  • Prior FDG-PET studies employing VBA that did not report hippocampal effects warrant re-examination using methods like HipMask.

Related Concept Videos