vol2Brain

vol2Brain performs automated dense anatomical segmentation and volumetric analysis of whole-brain magnetic resonance imaging (MRI) to quantify neuroanatomical structures and remain robust to white matter lesions (WML).

Key Features:

• Dense Anatomical Labeling: Labels over 100 distinct brain regions to provide high-granularity neuroanatomical segmentation.
• Robustness to White Matter Lesions (WML): Handles anatomical alterations such as WML to maintain segmentation accuracy in compromised brains.
• Multiscale Multi-Atlas Label Fusion: Implements a fast, multiscale multi-atlas label fusion approach that enables systematic error correction and precise volumetric information.
• Fully Automatic Pipeline: Executes segmentation and volumetric analysis without manual intervention.
• Evolution from volBrain: Extends the number of analyzable regions compared with the predecessor volBrain.

Scientific Applications:

• Longitudinal Studies: Provides detailed volumetric measures suitable for tracking structural changes over time.
• Disease Progression Monitoring: Supports monitoring of conditions such as multiple sclerosis, Alzheimer's disease, and other neurological disorders characterized by WML.
• Therapeutic Response Evaluation: Quantifies volumetric changes to assess treatment effects in clinical and research settings.

Methodology:

The pipeline combines multiscale multi-atlas label fusion and advanced deep learning techniques with systematic error correction to produce accurate volumetric brain segmentations.