**13. Conclusions**

In this paper, we have discussed some of the di fferences in current clinical MRI practices with optimized and UHF-MRI methods commonly employed in research environments. Clinical MRI hinges on weighted imaging with anisotropic voxel sizes and maintaining short acquisition, therefore being limited in signal and resolution. These current clinical practices are FDA-approved and are therefore deemed acceptable for neurosurgical purposes. However, the presence of side e ffects and non-responding patients nonetheless exist. Optimized 3 T and UHF-MRI tend to incorporate isotropic high-resolution imaging with quantitative and susceptibility-based contrasts for better visualization of deep brain structures, which, however, require more complex pre-processing and longer scan durations. The limitations incurred regarding reduced signal in clinical MRI and increased acquisition time with optimized 3 T can be largely overcome with the use of UHF-MRI. However, many of the image registration, correction, and post-processing techniques will typically require expertise that is outside the realm of traditional clinical settings. Importantly, the use of UHF-MRI and alternative contrasts such as QSM can only be implemented once pre-operative planning systems allow for their compatibility, which will require further FDA approval, not only for the MRI system but also for specific sequences. Additional approval for clinical use may be required for pre- and post-processing, such as the algorithms used for registration or calculation of quantitative maps.

We therefore propose that where UHF-MRI is not accessible, higher quality imaging can be obtained with optimized 3 T, although this will take longer than is perhaps clinically feasible for patients with severe movement disorders. Continued direct collaboration and combined e fforts between fundamental neuroscience researchers and clinicians will be essential for the development of optimized 3 T and UHF-MRI in the pre-operative planning process for DBS of the STN in PD. Multi-site clinical trials can facilitate the optimization and validation of certain sequences. Sequences with identical parameters should be compared on identical MRI systems and di fferent sites to ensure harmonization and reliability, as well as to validate the desired sequences. Rates of deviations between planned and actual target locations should be compared across vendors and systems as well as across sequences. Similarly, access agreements to work-in-progress protocols from MR vendors would facilitate the development and optimization of sequences, and would open access to underling algorithms and adjustable parameters within to pre-operative planning software vendors (e.g., Medtronic, St. Judes, Brainlab, Abott, Nextim, and Boston Scientific).

Of note, while this paper focused specifically on the STN as the most popular target for DBS in PD, alternative targets also exist (for example, see Figure 1). Some centers have long preferred the internal segmen<sup>t</sup> of the globus pallidus, and more recent research is being conducted on the suitability of alternate areas such as the ventral intermediate nucleus or the pedunculopontine nucleus for DBS targets. For a more in-depth review, please see [10,239,240] and the references therein.

**Author Contributions:** Conceptualization, B.R.I., M.C.K., A.A., Y.T., P.-L.B., and B.U.F.; validation, B.R.I., M.C.K., A.A., Y.T., P.-L.B., and B.U.F.; investigation, B.R.I., M.C.K., A.A., Y.T., P.-L.B., and B.U.F.; resources, Y.T. and B.U.F.; data curation, X.X.; writing—original draft preparation, B.R.I., M.C.K., and B.U.F.; writing—review and editing, B.R.I., M.C.K., A.A., Y.T., P.-L.B., and B.U.F.; visualization, B.R.I. and A.A.; supervision, B.U.F. and P.-L.B.; project administration, B.R.I. and M.C.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** The work was supported by a Vici gran<sup>t</sup> by the Dutch Organization for Scientific Research (NWO 016.Vici.185.052) (B.U.F.) and a starter gran<sup>t</sup> from the European Research Council (ERC-Stg 313481) (B.U.F.).

**Conflicts of Interest:** The authors declare no conflict of interest.
