A semi-automated brain atlas-based analysis pipeline for c-Fos immunohistochemical data

J Neurosci Methods. 2021 Jan 15:348:108982. doi: 10.1016/j.jneumeth.2020.108982. Epub 2020 Oct 20.


Background: The use of immunohistochemistry to quantify neural markers in various brain regions is a staple of neuroscience research. Numerous programs exist to automate quantification, but manual assignment of regions of interest (ROIs) within individual brain sections remains time consuming and can introduce interobserver variability.

New method: We have developed a novel open source FIJI-based immunohistochemical data analysis pipeline, Atlas-Based Analysis (ABA). ABA uses landmark-based image warping to adjust the experimental image to closely align with a published rat brain atlas. c-Fos positive cells are then quantified within predetermined ROI coordinates derived from the brain atlas. Image warping adjusts for natural variation in brain sections to ensure reliable alignment of ROIs for data analysis. This pipeline can be adapted for new atlases, landmarks, ROIs, and quantification measurements.

Results: ABA permits rapid quantification of immunoreactivity in multiple ROIs and produces results with high levels of interobserver consistency.

Comparison with existing methods: Compared to manual ROI designation, ABA reduces total analysis time by ∼70%. With correct use of landmarks for image warping, ABA produces similar results to manually drawn ROIs, results in no interobserver variability, and maintains c-Fos+ pixel dimensions.

Conclusions: ABA reduces time to obtain reliable results when performing automated immunoreactivity quantification and allows multiple users to analyze data without compromising the reliability of data obtained.

Keywords: Image warping; Immunohistochemistry; Rat brain atlas; c-Fos.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Brain Mapping*
  • Brain* / diagnostic imaging
  • Histological Techniques
  • Humans
  • Image Processing, Computer-Assisted
  • Magnetic Resonance Imaging
  • Observer Variation
  • Reproducibility of Results