Co‐registration of in vivo human MRI brain images to postmortem histological microscopic images

Abstract

Certain features such as small vascular lesions seen in human MRI are detected reliably only in postmortem histological samples by microscopic imaging. Co-registration of these microscopically detected features to their corresponding locations in the in-vivo images would be of great benefit to understanding the MRI signatures of specific diseases. Using non-linear Polynomial transformation, we report a method to co-register in-vivo MRIs to microscopic images of histological samples drawn off the postmortem brain. The approach utilizes digital photographs of postmortem slices as an intermediate reference to co-register the MRIs to microscopy. The overall procedure is challenging due to gross structural deformations in the postmortem brain during extraction and subsequent distortions in the histological preparations. Hemispheres of the brain were co-registered separately to mitigate these effects. Approaches relying on matching single-slices, multiple-slices and entire volumes in conjunction with different similarity measures suggested that using four slices at a time in combination with two sequential measures, Pearson correlation coefficient followed by mutual information, produced the best MRI-postmortem co-registration according to a voxel mismatch count. The accuracy of the overall registration was evaluated by measuring the 3D Euclidean distance between the locations of microscopically identified lesions on postmortem slices and their MRIpostmortem co-registered locations. The results show a mean 3D displacement of 5.1 ± 2.0 mm between the in-vivo MRI and microscopically determined locations for 21 vascular lesions in 11 subjects.

Registration of in-vivo to ex-vivo MRI of surgically resected specimens: A pipeline for histology to in-vivo registration

“…Many previous studies in MRI and histology registration (see Table 3 2004; Meyer et al, 2006;Lebenberg et al, 2010;Schormann et al, 1995;Yelnik et al, 2007;Osechinskiy and Kruggel, 2011;Bardinet et al, 2002), and furthermore many previous studies included evaluation on only one dataset (Malandain et al, 2004;Choe et al, 2011;Meyer et al, 2006;Schormann et al, 1995;Kim et al, 2000;Yelnik et al, 2007;Osechinskiy and Kruggel, 2011;Bardinet et al, 2002;Lazebnik et al, 2003). Of the studies that did report accuracy on more than one dataset, TRE ranged from sub-millimeter (Ceritoglu et al, 2010;Jacobs et al, 1999;Yang et al, 2012) to 3-5 mm (Liu et al, 2012;Singh et al, 2008). Techniques that reported sub-millimeter TRE were applied on either whole brain sections of rodents or serially sectioned histology of primates, thus the smaller scale of anatomy and lack of variable resection boundaries can explain the lower TRE relative to our method.…”

“…Techniques that reported sub-millimeter TRE were applied on either whole brain sections of rodents or serially sectioned histology of primates, thus the smaller scale of anatomy and lack of variable resection boundaries can explain the lower TRE relative to our method. For a more relevant comparison (Singh et al, 2008), performed registration of human in-vivo and post-mortem whole brain specimens and reported a TRE of 5.1 mm. The only existing work that dealt with resected temporal lobe specimens was (Eriksson et al, 2005), however they only aimed to find corresponding slices between MRI and histology, and reported inter-and intra-observer variability (<2 mm) instead of an accuracy measure.…”

“…The majority of these studies focused on primates (Dauguet et al, 2007;Malandain et al, 2004;Breen et al, 2005;Ceritoglu et al, 2010;Choe et al, 2011) or rodents (Jacobs et al, 1999;Humm et al, 2003;Meyer et al, 2006;Lebenberg et al, 2010;Yang et al, 2012;Liu et al, 2012). The few studies that registered human brain MRIto histology were performed on wholebrain (Schormann et al, 1995;Kim et al, 2000;Singh et al, 2008), or single hemisphere (Yelnik et al, 2007;Osechinskiy and Kruggel, 2011) post-mortem serially sectioned data (Amunts et al, 2013) created a 3D model of single subject's brain using post-mortem histological sections reconstructed at 20 m isotropic resolution and registered it to a T1 average atlas created from 24 subjects. Eriksson et al (2005) reported registering histology of neocortical specimens from anterior temporal lobectomies to in-vivo MRI; however, their approach only involved visually selecting the closest coronal MRI slice for each histology slide, and did not attempt to find a dense correspondence between each histology slide and its corresponding MRI slice.…”

Deformable image registration between pathological images and MR image via an optical macro image

“…Among these studies, several developed methods can be used for images of the brains of small animals, such as rats, mice, and rabbits [15,16,25,29], while others employed methods may be applied to a small part of the human brain [11,22]. Ideally, a whole human organ should be analyzed with the PT and MR images.…”

High-Resolution Fiber Tract Reconstruction in the Human Brain by Means of Three-Dimensional Polarized Light Imaging

“…Such principle of measurement is referred to as polarimetry and has been used in anatomical studies of the central nervous system already a century ago (Brodmann, 1903). However, in the recent past, significant advances in the 3D reconstruction of microtome sections (Dauguet et al, 2007; Singh et al, 2008; Capek et al, 2009; Palm et al, 2010), image analysis, computational techniques, and progress in understanding the interaction of polarized light with birefringent tissue (Schnabel, 1966; Fraher and MacConnaill, 1970; Oldenbourg and Mei, 1995;Oldenbourg, 1996; Oldenbourg et al, 1998; Massoumian et al, 2003; Farrell et al, 2005; Larsen et al, 2007; Axer et al, 2011) have opened up new avenues to study brain regions with complex fiber architecture at the highest level of detail. We took advantage of this progress to gain a vector field description of fiber tract orientations in histological brain sections and to reconstruct 3D fiber tract models in selected brain regions across a series of aligned sections.…”