An information theory model for optimizing quantitative magnetic resonance imaging acquisitions.

Area of Science:

• Medical Imaging
• Information Theory
• Quantitative MRI

Background:

• Quantitative MRI (qMRI) acquisition parameter selection is often empirical, involving trial and error.
• Optimizing parameters is crucial for minimizing variability in quantitative maps and synthetic image generation.
• Current methods lack a quantitative approach for informed parameter selection.

Purpose of the Study:

• To introduce and evaluate a quantitative method for selecting qMRI acquisition parameters that minimize image variability.
• To apply an information theory framework using mutual information for parameter optimization.
• To assess the effectiveness of the proposed method on artificial, phantom, and in vivo data.

Main Methods:

• Developed an information theory framework utilizing mutual information to quantify information content of potential acquisitions.
• Applied the framework to a 3D-quantification using an interleaved Look-Locker acquisition sequence with T2 preparation pulse (3D-QALAS) signal model.
• Validated the method using artificial data, a multiparametric imaging standard phantom, and in vivo human brain measurements.

Main Results:

• Higher mutual information calculated by the model correlated with a smaller coefficient of variation in reconstructed parametric maps from phantom data.
• Information-based calibration of acquisition parameters in vivo demonstrated a decrease in parametric map variability.
• Results were consistent with the model's predictions, showing improved quantitative map consistency.

Conclusions:

• The developed information theory framework provides a quantitative method for optimizing qMRI acquisition parameters.
• This approach effectively minimizes image variability, leading to more reliable quantitative MRI data.
• Information-based parameter selection offers a significant improvement over empirical methods in qMRI.