Methodology for computing white matter nerve fiber orientation in human histological slices

Area of Science:

• Neuroscience
• Medical Imaging
• Computational Biology

Background:

• Histological analysis of brain biopsies is the gold standard for mapping white matter nerve fiber orientation.
• Accurate nerve fiber mapping is essential for validating non-invasive techniques like diffusion MRI.
• Manual extraction of nerve fiber directions from histological slices is labor-intensive and error-prone.

Purpose of the Study:

• To develop and validate a semi-automated algorithm for estimating nerve fiber directionality in human brain white matter.
• To reduce the time and potential for human error associated with manual histological analysis.
• To provide a reliable method for generating orientation data to support diffusion MRI validation.

Main Methods:

• A semi-automated algorithm was developed to process histological images.
• The algorithm generates a binary-segmented mask of nerve fibers.
• Fourier-domain analysis is used to estimate average nerve fiber directionality and its uncertainty.

Main Results:

• The algorithm's orientation estimates showed good agreement with expert visual inspection (weighted RMS difference of 15.4°, improving to 9.9° with high expert confidence).
• Comparison with manual segmentation yielded an RMS difference of 11.2°.
• The method provides an uncertainty level for its directionality estimates.

Conclusions:

• The semi-automated method demonstrates good agreement with established manual expert-based approaches.
• This algorithm offers a more efficient and potentially more accurate alternative for analyzing histological slices.
• The findings support the use of this semi-automated technique for research in human brain white matter.