Robust thalamic nuclei segmentation from T1-weighted MRI using polynomial intensity transformation.

Area of Science:

• Neuroimaging
• Medical Image Analysis
• Brain Anatomy

Background:

• Accurate segmentation of thalamic nuclei is vital for understanding cognition and neurological diseases.
• Standard T1-weighted MRI offers poor contrast for detailed thalamic segmentation.
• White-matter-nulled (WMn) MRI improves contrast but is not widely available.

Purpose of the Study:

• To introduce Histogram-based Polynomial Synthesis (HIPS) for synthesizing WMn-like contrast from standard T1w MRI.
• To integrate HIPS into the THOMAS segmentation pipeline (HIPS-THOMAS).
• To evaluate HIPS-THOMAS against other segmentation methods across various MRI parameters.

Main Methods:

• Developed HIPS, a preprocessing technique using polynomial approximation for intensity transformation.
• Integrated HIPS into the THalamus Optimized Multi-Atlas Segmentation (THOMAS) pipeline.
• Compared HIPS-THOMAS with a CNN-based method and T1w-THOMAS across diverse MRI contrasts, manufacturers, and field strengths.

Main Results:

• HIPS significantly improved intra-thalamic contrast and boundary definition on T1w MRI.
• HIPS-THOMAS achieved higher Dice coefficients and lower volume errors than competing methods.
• HIPS demonstrated the least inter-scanner variability in phantom studies.

Conclusions:

• HIPS is an effective and robust method for enhancing thalamic nuclei segmentation using standard T1w MRI.
• The HIPS-THOMAS pipeline offers improved accuracy and reliability for thalamic segmentation.
• This technique can broaden the accessibility of detailed thalamic analysis in clinical and research settings.