Josefina Maranzano

I chose this topic because I considered it essential to assess the potential of 3T multicontrast cortical lesion detection protocols. Cortical lesions represent a large percentage of the overall lesion load in multiple sclerosis. Unfortunately, they are more difficult to detect in clinical standard 1.5T and 3T MRI scans. 7T scanners have given very promising results in detecting cortical lesions, but these scanners are not widely available, so the majority of people with multiple sclerosis will not be assessed using 7T scanners in the near future.

Fortunately, current image processing tools allow the accurate coregistration of multiple contrasts acquired at a reasonably high resolution (1-mm isotropic voxel T1-weighted and FLAIR contrasts) using 3T scanners. These contrasts, coregistered with T2-weighted and proton-density-weighted contrasts, may provide cortical lesion counts closer in number to those provided by 7T MRI scans. Because 7T scanners are not widely available for clinical or extensive research studies, I considered it important to perform a more in-depth assessment of multicontrast 3T reading protocols as tools to detect cortical lesions.

This article shows that leukocortical lesions (type I) are detected well by 3T multicontrast identification protocols. However, 7T scans are more sensitive to subpial lesions (types III and IV), which are only very partially captured by 3T scans (low sensitivity of 3T [11%] in relation to 7T for subpial demyelination). Interestingly, the overall number of cortical lesions detected by 3T scans is significantly associated with the number of subpial lesions detected by 7T scans, so 3T multicontrast cortical lesion counts could be viewed as an accessible tool to roughly assess overall cortical pathology in multiple sclerosis.

I am currently working on assessing the dynamic changes of the different types of cortical lesions in subsequent scans of the patients in this study (scans acquired 2 years later) using 7T scans because, as mentioned before, they are more sensitive to assess subpial demyelination. The manual detection of different cortical lesion types on 7T scans using multiple contrasts is a time-consuming task, but I am hopeful that I will have some results to share with the public by the end of this year.

Read this article at AJNR.org …