Reporting of quantitative oxygen mapping in EPR imaging

Abstract

Oxygen maps derived from electron paramagnetic resonance spectral-spatial imaging (EPRI) are based upon the relaxivity of molecular oxygen with paramagnetic spin probes. This technique can be combined with MRI to facilitate mapping of pO(2) values in specific anatomic locations with high precision. The co-registration procedure, which matches the physical and digital dimensions of EPR and MR images, may present the pO(2) map at the higher MRI resolution, exaggerating the spatial resolution of oxygen, making it difficult to precisely distinguish hypoxic regions from normoxic regions. The latter distinction is critical in monitoring the treatment of cancer by radiation and chemotherapy, since it is well-established that hypoxic regions are three or four times more resistant to treatment compared to normoxic regions. The aim of this article is to describe pO(2) maps based on the intrinsic resolution of EPRI. A spectral parameter that affects the intrinsic spatial resolution of EPRI is the full width at half maximum (FWHM) height of the gradient-free EPR absorption line in frequency-encoded imaging. In single point imaging too, the transverse relaxation times (T(2)(∗)) limit the resolution since the signal decays by exp(-t(p)/T(2)(∗)) where the delay time after excitation pulse, t(p), is related to the resolution. Although the spin densities of two point objects may be resolved at this separation, it is inadequate to evaluate quantitative changes of pO(2) levels since the linewidths are proportionately affected by pO(2). A spatial separation of at least twice this resolution is necessary to correctly identify a change in pO(2) level. In addition, the pO(2) values are blurred by uncertainties arising from spectral dimensions. Blurring due to noise and low resolution modulates the pO(2) levels at the boundaries of hypoxic and normoxic regions resulting in higher apparent pO(2) levels in hypoxic regions. Therefore, specification of intrinsic resolution and pO(2) uncertainties are necessary to interpret digitally processed pO(2) illustrations.

Fig. 1

Schematic of a resolution phantom consisting of a hollow Lucite cylinder with rods, tubes and rectangular slabs positioned inside (shaded regions) such that when filled with a spin probe, transverse 2D view perpendicular to the cylinder axis appears as on right with shaded portions being filled with the spin probe and the white portions being spin-free.

Fig. 2

The image of the phantom in Fig. 1 at different resolutions. A slice of MR image at size 256 × 256 was used to calculate the low-resolution images. The object size appears to be artificially larger at low resolution.

Fig. 3

Digital enhancement of a low resolution image. A 2D spin density image of the phantom filled with Oxo-63 spin label. Images were reconstructed at digital resolutions: (a) at intrinsic resolution (n = 61), (b) reconstructed at image size of 128 × 128, and (c) reconstructed at image size 256 × 256. The circular shapes appear better in (c) than (a) though the intrinsic resolution is the same for all the three images.

Fig. 4

Comparison of intrinsic resolution with digital enhancement of a 2D spin density image of a phantom filled with Oxo-63 spin label. A. MRI scanned at 0.125 mm resolution. B, C, and D are EPR images scanned at the intrinsic resolutions of 0.42, 1.21 and 1.42 mm respectively. The horizontal gray lines in image indicate a row selected for signal intensity profile shown below the image. E - H: Down sized images of A – C to match with matrix size of D. I – L:The images A – D reconstructed at n=256 for co registration. Note the blur in EPR images G and H because of low intrinsic resolution in spite of equal digital resolution with F.

Fig. 5

Coregistration of EPR spin density image with MR anatomic image of a tumor bearing mouse hind leg. (a) MRI at resolution 0.109 mm. (b) Solid lines drawn at the intensity gradients trace the anatomic shape. (c) EPR image at its intrinsic resolution of 1.55 mm and overlay of the edges from MRI. High resolution MRI lends support to low resolution EPRI to delineate anatomic locations. Data are from a slice of 3D image.

Fig. 6

Resolution indication on the image. (a) pO₂ map at intrinsic resolution of 1.55 mm. (b) Overlay of the anatomic shape from MRI. (c) Digitally enhanced to match to MRI size of 256 × 256. The spatial resolution is shown on spatial dimension. The uncertainty level of pO₂ is shown on pO₂ scale.

Fig. 7

Simulation of signals using hypothetical point objects having spin densities of 6, 6, 6, 1, 1. The objects are equally spaced. (a) Spacing = Δ. Grid lines indicate the pixels boundaries at this spacing. (b) Spacing = 2Δ. (c) Effect of line width variation on (a). The line widths of objects (i) and (ii) are 1.5Δ and 2Δ respectively. Line width increase due to pO₂ leads to higher overlaps. (d) Signal decay with time in SPI for hypothetical T₂ values. Relative signal levels with time are shown at T₂ values representing the line widths of Δ, 1.5Δ (i) and 2Δ (ii) respectively. The horizontal line at 1% of the signal intersecting the decay curves at different t_p values indicates the limits available for SPI at a hypothetical SNR level. SPI resolutions available for pO₂ mapping are lower than oxygen free conditions because of reduced t_p limits.

Fig. 8

(a) A hypothetical pO₂ image schematically representing tumor (left) with hypoxic core (0%), surrounding high oxygen region (5%) and normoxic regions (4%). The two dots represent fiducials. The thin white lines indicate boundaries of pO₂ regions. (b) The point spread function (Gaussian noise) (c) Convolution of (a) and (b). The blurring leads to distorted pO₂ values at the region boundaries. Note that the fiducials that are at normoxic levels are highly diffused by noise to almost hypoxic levels. At region boundaries hypoxic regions appear to have higher pO₂ values (0 – 2.5%) while normoxic regions appear to be at lower pO₂ levels (2 – 4%). (d) pO₂ profile across the tumor indicated by dashed line is shown below (c). Blurred image profile appears as a curve tracking pO₂ levels before blurring.

Fig. 9

Spin density 3D EPR image of a tumor bearing mouse leg. (a) The positioning of the tumor-bearing leg on a 17 mm resonator. Surface-rendered images at identical angles and cut-off intensity (b) before and (c) after deconvolution.

Fig. 10

pO₂ map of a slice from 3D EPR image of a tumor bearing mouse leg at the intrinsic resolution of 1.55 mm. (a) Before deconvolution. (b) pO₂ map calculated using spin densities after deconvolution as shown in Fig. 9b. (c) pO₂ map obtained by the deconvolution of the one derived from raw images (9a).