Meyer's loop tractography for image-guided surgery depends on imaging protocol and hardware

• Published in: NeuroImage clinical Vol. 20; pp. 458 - 465
• Main Authors: Chamberland, Maxime, Tax, Chantal M.W., Jones, Derek K.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier Inc 01.01.2018; Elsevier
• Subjects: Computers; Databases, Factual; Diffusion Tensor Imaging - methods; Epilepsy, Temporal Lobe - diagnostic imaging; Epilepsy, Temporal Lobe - physiopathology; Epilepsy, Temporal Lobe - surgery; Humans; Radiology; Regular; Surgery, Computer-Assisted - methods; Temporal Lobe - diagnostic imaging; Temporal Lobe - physiopathology; Temporal Lobe - surgery; Visual Field Tests - methods; Visual Fields - physiology; Visual Pathways - diagnostic imaging; Visual Pathways - physiopathology; Visual Pathways - surgery
• DOI: 10.1016/j.nicl.2018.08.021
• ISSN: 2213-1582

Surgical resection is an effective treatment for temporal lobe epilepsy but can result in visual field defects. This could be minimized if surgeons knew the exact location of the anterior part of the optic radiation (OR), the Meyer's loop. To this end, there is increasing prevalence of image-guided surgery using diffusion MRI tractography. Despite considerable effort in developing analysis methods, a wide discrepancy in Meyer's loop reconstructions is observed in the literature. Moreover, the impact of differences in image acquisition on Meyer's loop tractography remains unclear. Here, while employing the same state-of-the-art analysis protocol, we explored the extent to which variance in data acquisition leads to variance in OR reconstruction. Diffusion MRI data were acquired for the same thirteen healthy subjects using standard and state-of-the-art protocols on three scanners with different maximum gradient amplitudes (MGA): Siemens Connectom (MGA = 300 mT/m); Siemens Prisma (MGA = 80 mT/m) and GE Excite-HD (MGA = 40 mT/m). Meyer's loop was reconstructed on all subjects and its distance to the temporal pole (ML-TP) was compared across protocols. A significant effect of data acquisition on the ML-TP distance was observed between protocols (p < .01 to 0.0001). The biggest inter-acquisition discrepancy for the same subject across different protocols was 16.5 mm (mean: 9.4 mm, range: 3.7–16.5 mm). We showed that variance in data acquisition leads to substantive variance in OR tractography. This has direct implications for neurosurgical planning, where part of the OR is at risk due to an under-estimation of its location using conventional acquisition protocols. [Display omitted] •Diffusion MRI data from 13 subjects on 3 scanners with different maximum gradient amplitude.•Optic radiation tractography was performed using “standard” and “state-of-the-art” acquisitions.•Meyer's loop to Temporal Pole distance varies up to 16.5 mm for the same subject across protocols.•State-of-the-art protocols provide Meyer's loop measurements that are in line with ex-vivo studies.•Surgical planning for anterior temporal lobe resection depends on protocol and available hardware.