Evaluation of hippocampal volume based on MR imaging in patients with bipolar affective disorder applying manual and automatic segmentation techniques

J Magn Reson Imaging. 2011 Mar;33(3):565-72. doi: 10.1002/jmri.22473.

Abstract

Purpose: To compare the hippocampal volumes in patients with bipolar disorder (BD) and healthy controls, obtained by applying different segmentation methods (manual, Freesurfer [FS], and FSL).

Materials and methods: The study included 27 patients with BD and 40 healthy controls. T1-weighted images in the sagittal plane were acquired on a 3 Tesla (T) MR scanner. Hippocampal volumetry was performed using one manual and two automated methods (FS and FSL). One-way repeated analysis of variance was applied to test the differences in hippocampal volumes using the three segmentation methods. To evaluate the agreement among the three tested volumetric segmentation methods the intraclass correlation coefficients (ICCs) were calculated.

Results: Hippocampal volumes obtained from all methods were significantly different (P < 0.05) in BD patients after intracranial volume correction, indicating a reduction in volume, unless from the manual method of the left hippocampal volume. The ICCs of the hippocampal volume between the manual method and FS were 0.846 (right) and 0.859 (left), and between the manual method and FSL were 0.746 (right) and 0.654 (left).

Conclusion: Both manual and automatic segmentation methods detected reductions in the hippocampal volumes in BD patients. Automated segmentation methods are a robust and reproducible option for assessing hippocampal volume.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Automation
  • Bipolar Disorder / pathology*
  • Brain Mapping / methods
  • Diagnostic Imaging / methods
  • Female
  • Hippocampus / pathology*
  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Magnetic Resonance Imaging / methods*
  • Male
  • Middle Aged
  • Organ Size
  • Reproducibility of Results