Direct magnetic resonance imaging of histological tissue samples at 3.0T

Abstract

Direct imaging of a histological slice is challenging. The vast difference in dimension between planar size and the thickness of histology slices would require an RF coil to produce a uniform RF magnetic (B1) field in a 2D plane with minimal thickness. In this work a novel RF coil designed specifically for imaging a histology slice was developed and tested. The experimental data demonstrate that the coil is highly sensitive and capable of producing a uniform B1 field distribution in a planar region of histological slides, allowing for the acquisition of high-resolution T2 images and T2 maps from a 60-microm-thick histological sample. The image intensity and T2 distributions were directly compared with histological staining of the relative iron concentration of the same slice. This work demonstrates the feasibility of using a microimaging histological coil to image thin slices of pathologically diseased tissue to obtain a precise one-to-one comparison between stained tissue and MR images.

FIG. 1

a: Schematic of the flat slotted-tube resonator. Two separate copper strips are connected by a capacitor at each of the four corners. b: Calculated transverse B₁ field component perpendicular to B₀ in the center plane of the coil. The imaging region where the glass coverslips are placed is outlined by the white dashed box. The profiles of the B₁ field along the cross hairs through the imaging region are shown above and to the right of the B₁ map. The direction of the B₀ field is noted to aid in the orientation of the coil within the magnet.

FIG. 2

a: Schematic of the optimized histological coil. A continuous strip of copper is wrapped around the glass coverslips and a piece of Teflon (green). Two glass coverslips with the histological sample encased between them are inserted into the opening of the coil for tissue loading. b: Calculated transverse B₁ field perpendicular to B₀ in the center plane of the histological coil. The imaging region where the glass coverslips are placed is enclosed in the white dashed line. The profiles along the cross hairs demonstrate the homogeneity of the coil. The red outlined box represents the region where the Teflon is placed. The color scales in Fig. 1b and 2b are set the same to facilitate comparison between the two designs. The profiles in Fig. 1b and 2b indicate that the homogeneity of the B₁ field is improved with the optimized histological coil design. The direction of the B₀ field is noted to aid in the orientation of the coil within the magnetic field.

FIG. 3

a: Computer model of the histological coil. The Delrin acetal resin where the coil is positioned is represented by the semitransparent object. b: Photograph of the completed coil with the 370-μm opening on the side where the tissue sample and coverslips are inserted along with tuning and matching variable capacitors and BNC connector at the end.

FIG. 4

a: T₂-weighted image obtained with the histological coil of the 12.5-μl phantom approximating the size of a tissue sample. Profiles through the horizontal and vertical center of the phantom are presented on the top and to the side of the image. b: B₁-field map of the 43.2-μl phantom with profiles demonstrating the B₁-field uniformity of the coil within the imaging region. There are signal dropouts at the bottom two corners due to the drying of the agar sample, and two small artifacts due to air bubbles within the sample.

FIG. 5

a: T₂-weighted image of a 60-μm-thick slice of human brain tissue. b: Image of the same tissue section stained after MRI with modified Perl’s Prussian blue for iron content. Dark regions are indicative of higher iron concentration within the tissue. The correlation of iron content with T₂ contrast is shown in the images (the WM shows up as darker on the T₂ image and high in iron with the histological stain; conversely, GM regions show up as brighter on the T₂ image and lower in stained iron concentration). The 50× digital enlargement of the MR image and 50× optical bright-field magnification at the bottom demonstrate the close relationship between iron deposition and T₂ content. The regions of highly focal cellular iron concentration in the magnified histological image are clearly seen as darker spots in the MR image. The relative size of the iron deposition affects the magnitude of the hypointensities in the MR image. A typical arrowhead-type imaging artifact is visible around the large hypointensity in the digitally enlarged MR image. Staining artifacts at the edges of the tissue sample due to increased edge surface area are not indicative of higher iron content. Scale bars of 2000 μm for the whole-tissue images and 250 μm for the 50× magnified images are included for reference. The direction of the B₀ field is noted to aid in the orientation of the slice within the histological coil.

FIG. 6

a: T₂ parameter map generated from the same histological slice data shown in Fig. 5a. ROIs were selected in GM (solid box) and WM (dashed box) from six individual maps. b: Average T₂ relaxation curves (N = 6) with SD bars of each echo image from the ROIs. The average T₂ in the ROI is 85.5 ± 2.4 ms for GM and 66.1 ± 1.5 ms for WM. The direction of the B₀ field is noted to aid in the orientation of the slice within the histological coil.