Spatial normalization of ultrahigh resolution 7 T magnetic resonance imaging data of the postmortem human subthalamic nucleus: a multistage approach

Max C Keuken,Stefan Geyer,Anneke Alkemade,Birte U Forstmann,Christa Műller-Axt,Marcel Weiss,Robert Turner

doi:10.1007/s00429-014-0754-4

Abstract

In this paper, we describe a novel processing strategy for the spatial normalization of ultrahigh resolution magnetic resonance imaging (MRI) data of small ex vivo samples into MNI standard space. We present a multistage scanning and registration method for data of the subthalamic nucleus (STN) obtained using ultrahigh 7 T MRI on four human postmortem brain samples. Four whole brains were obtained and subjected to multistage MRI scanning, corresponding to four different brain dissection stages. Data sets were acquired with an isotropic resolution of 100 μm enabling accurate manual segmentation of the STN. Spatial normalization to MNI reference space was performed, probability maps were calculated, and results were cross-checked with an independent in vivo dataset showing significant overlay. Normalization of results obtained from small tissue samples into MNI standard space will facilitate comparison between individual subjects, as well as between studies. When combining ultrahigh resolution MRI of ex vivo samples with histological studies via blockface imaging, our method enables further insight and inference as multimodal data can be compared within the same reference space. This novel technique may be of value for research purposes using functional MRI techniques, and in the future may be of assistance for anatomical orientation in clinical practice.

Highlights

Recent studies have recognized the importance of the integration of data available from histological analyses with noninvasive imaging techniques (Annese 2012; Yang et al 2013)
In this paper, we describe a novel processing strategy for the spatial normalization of ultrahigh resolution magnetic resonance imaging (MRI) data of small ex vivo samples into MNI standard space
The high anatomical detail is likely to have contributed to the differences observed in subthalamic nucleus (STN) volume (109 ± 46 mm3), which we reported earlier (Keuken et al 2013)

Summary

Introduction

Recent studies have recognized the importance of the integration of data available from histological analyses with noninvasive imaging techniques (Annese 2012; Yang et al 2013). Histological analyses have been used to provide impressive three-dimensional (3D) reconstructions of the human brain (Amunts et al 2013) These studies are valuable, but the techniques used may not be universally suitable for routine implementation (Annese 2012). This is illustrated by the low number of observations described in these studies, n = 1, 1 and 2 respectively (Annese 2012; Amunts et al 2013; Yang et al 2013). The need for methods allowing the translation of histological data from small tissue samples into standard MNI space is evident. The applicability of the created pipeline is illustrated using the human subthalamic nucleus (STN)

Objectives

Methods

Results

Conclusion