Rejoinder: LESA: Longitudinal Elastic Shape Analysis of Brain Subcortical Structures

The brain surfaces including both cortical and subcortical structures including hippocampus have been analyzed for more than a decade using publicly available software packages such as FreeSurfer (Dale & Fischl 1999) and SurfStat (Worsley et al. 2009, Chung et al. 2010). Zhang et al. (2022) proposes an elastic shape metric based method for perform-ing longitudinal shape analysis on brain subcortical structures. However, the demonstrated applications are limited to global summary measures such as the total surface area and prin-ciple component (PC) scores significantly limiting the impact of the study. For analyzing total surface area, we do not even need to align structures using LESA. PC scores loose richer vertex-based local information and it is unclear what parts of the hippocampus are responsible for longitudinal change. A more effective approach is to perform local shape analysis using the deformation-based morphometry (DBM) and tensor-based morphometry (TBM) after obtaining deformation in LESA (Ashburner et al. 1998, 2000, Thompson et al. 2000). Considering elastic methods put severe constraints on the Jacobian determinant of image deformation (Chung et al. 2001), it is not clear LESA can be effectively used in local shape analysis. We contrast shape analysis done in Zhang et al. (2022) against DBM and TBM in a longitudinal hippocampus study (Chung et al. 2011). The deformation-based morphometry (DBM) utilizes the deformation field obtained