Overestimation of cerebral aneurysm wall thickness by black blood MRI?

Recently, Park et al (1) reported unruptured saccular cerebral aneurysm (USCA) wall thicknesses on the order of 0.5 mm using a black blood double-inversion (BBDI) protocol having an in-plane resolution of 0.5 mm and slice thickness of 3 mm. Early in their discussion of the results the authors concede that their measurements were greater than the “20–500 micron” (p. 1181) wall thicknesses reported by specimen-based studies and attribute this to “different state of the aneurysms between [their] study (in vivo and unruptured state) and previous studies (ex vivo and ruptured state) (p. 1182).” Later, in discussing the potential limitations of their study, they add that their measured wall thicknesses were “similar to the pixel size of the BBDI; therefore, there might have been measurement errors (p. 1183).” In fact, there is a good chance those measurement errors, rather than differences in the states of the aneurysms, are at the root of the discrepancy between the MRI- and specimen-based thickness measurements. Referring to fig. 8 of our recent study on wall thickness (2) overestimation, an aneurysmal wall having a true thickness of, say, 250 μm would, at 0.5 mm acquired resolution, appear to have a thickness closer to 0.7 mm. That the authors report 0.5–0.6 mm wall thickness suggests that ours is likely the worst-case scenario, since it ignores the potentially salutary influence of surrounding tissue, and of operator judgment in manual segmentation vs. our automated edge detection. Nevertheless, it seems likely that the BBDI protocol used by the authors has served to overestimate the USCA wall thickness by more than they imply. Moreover, as we (2) and others (3) have also shown, such overestimation can be exacerbated by thick slice acquisitions when anatomical complexity and wall curvature may introduce obliqueness between the wall and slice plane, such as might be the case for smaller aneurysms. This effect only gets worse if slice thickness is sacrificed for higher in-plane resolution. As such, the authors are correct to conclude, “near isotropic resolution with 1024 matrices may be necessary for evaluation of the entire USCA wall (p. 1183).” However, what they fail to mention is that this will require at least an order-of-magnitude reduction in voxel volume compared to their current acquisition. While not wishing to discourage the noninvasive measurement of cerebral aneurysm wall thickness, we feel that a stronger message regarding the potential for severe overestimation due to partial volume effects is warranted in light of the spatial resolutions currently achievable, even at 3 Tesla (4). We do encourage the authors and others to carry out ex vivo imaging studies to demonstrate the true capabilities and limitations of MRI for resolving the walls of USCA. David A. Steinman PhD [email protected]*, Luca Antiga PhD , Bruce A. Wasserman MD , * Biomedical Simulation Laboratory Department of Mechanical & Industrial Engineering University of Toronto Toronto, ON, Canada, Biomedical Engineering Department Mario Negri Institute for Pharmacological Research Ranica (BG), Italy, Department of Radiology Johns Hopkins School of Medicine Baltimore, Maryland, USA.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.