Measuring T1 Relaxation Times at High Resolution from Rapid Brain MRI

<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> -weighted imaging is extensively employed in clinical MRI to visualize brain anatomy. Typically, these sequences are qualitative, serving primarily for visual interpretation. The <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> relaxation time, a quantitative metric of tissue characteristics, provides contrast in these images, thereby enhancing the visibility of structures based on their specific <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> properties, which in turn improves diagnostic precision. Accurately measuring <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> relaxation time necessitates multiple scans at varying time points following the application of radiofrequency pulses. Unfortunately, the time required for these acquisitions is impractically long for routine clinical MRI. This study seeks to establish a method for estimating high-resolution <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> maps from rapid brain MRI within a clinically acceptable duration. We devised an imaging protocol to acquire a minimal set of low-resolution <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> -weighted images in approximately 6 minutes. We developed a technique that integrates super-resolution and data synthesis strategies to generate the necessary number of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> -weighted images, allowing for voxel-wise estimation of a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> map at 0.7mm high isotropic resolution. We assessed our approach using a dataset that contains 90 brain MRI volumes from 10 subjects scanned with a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$3\mathrm{T}$</tex> MRI system. Comprehensive experiments demonstrated that our method surpassed several existing techniques in the performance of quantitative <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{T}_{1}$</tex> map estimation.