Search

We analyzed image characteristics in T(1)-, T(2)-, and diffusion-weighted in vitro magnetic resonance (MR) images acquired at predefined stages of the ovarian cycle in 36 heifers to test the hypothesis that MR image attributes of the follicle wall reflect the physiologic status of ovarian follicles (viable, atretic, dominant, subordinate). Numerical pixel values (NPV), standard deviation of pixel values (heterogeneity), and area under the curve were used to assess images of follicle walls. Pixel values of the wall were used to calculate a regression line from which intercept, slope, and coefficient of determination were calculated. In T(1) images, NPV of dominant follicles were less likely to fit a regression line at the preovulatory phase than at any other phase (P < 0.1). Preovulatory dominant follicles had lower area under the curve in diffusion-weighted images than early and late static dominant follicles of the anovulatory wave (P < 0.02). Subordinate follicles in the presence of a preovulatory dominant follicle had lower mean NPV in T(1)- and T(2)-weighted images and lower intercepts in T(1)-weighted images than subordinate follicles of the anovulatory wave (P < 0.02). Early atresia of dominant follicles was identified at the late static phase by greater area, mean NPV, and slope in T(2)-weighted images (P < 0.02). Preovulatory dominant follicles had poor fit of NPV to a regression line in T(1)-weighted images and lower area under the curve in diffusion images. Atretic follicles had brighter walls with more acute transitions from follicular fluid to stroma in T(2)-weighted images and more heterogeneous walls in diffusion images. The MR image attributes of the follicle wall reflected the physiologic status of dominant and largest subordinate follicles.

A robust standardized method for segmentation, quantification, and normalization of pediatric hippocampal volumes using magnetic resonance imaging is presented. The method will find application in time course measurements of hippocampal volumes in pediatric patients who suffer from temporal lobe epilepsy and was tested prospectively on six control patients (13-60 mo of age). The un-normalized hippocampal volumes obtained using our segmentation method ranged from 3.85 to 6.38 mL, in agreement with previously published results. Inter- and intraobserver variability of the segmentation method was determined to be 13.3% and 2.8%, respectively. Four different methods of volume normalization were tested. Normalization is required to adjust for age-related increases in hippocampal volume. The normalization approach that seemed to compensate best for growth-related hippocampal volume changes was based on a simple estimation of intracranial volumes. This is the first report of a consistent and reliable method for segmentation and normalization of hippocampi from pediatric patients that can be used to study the progression of neurologic diseases in children.

An algorithm of Dutt and Rokhlin (SIAM J Sci Comput 1993;14:1368-1383) for the computation of a fast Fourier transform (FFT) of nonuniformly-spaced data samples has been extended to two dimensions for application to MRI image reconstruction. The 2D nonuniform or generalized FFT (GFFT) was applied to the reconstruction of simulated MRI data collected on radially oriented sinusoidal excursions in k-space (ROSE) and spiral k-space trajectories. The GFFT was compared to conventional Kaiser-Bessel kernel convolution regridding reconstruction in terms of image reconstruction quality and speed of computation. Images reconstructed with the GFFT were similar in quality to the Kaiser-Bessel kernel reconstructions for 256(2) pixel image reconstructions, and were more accurate for smaller 64(2) pixel image reconstructions. Close inspection of the GFFT reveals it to be equivalent to a convolution regridding method with a Gaussian kernel. The Gaussian kernel had been dismissed in earlier literature as nonoptimal compared to the Kaiser-Bessel kernel, but a theorem for the GFFT, bounding the approximation error, and the results of the numerical experiments presented here show that this dismissal was based on a nonoptimal selection of Gaussian function.

The magnetic resonance images and maps of bovine ovaries acquired at defined phases of follicular development and regression were studied to determine whether magnetic resonance image attributes of the follicular antrum reflect the physiological status of dominant and subordinate ovarian follicles. Ovariectomies were performed at day 3 of wave one, day 6 of wave one, day 1 of wave two and at >/= day 17 after ovulation. The timings of ovariectomies were selected to acquire growing, early static, late static and regressing follicles of the first wave and preovulatory follicles of the ovulatory wave. Pre-selection and subordinate follicles were also available for analysis. Serum samples were taken on the day of ovariectomy and follicular fluid samples were taken after imaging. Numerical pixel value and pixel heterogeneity in a spot representing approximately 95% of the follicular antrum were quantified in T(1)- and T(2)-weighted images. T(1) and T(2) relaxation rates (T(1) and T(2)), proton density, apparent diffusion coefficients and their heterogeneities were determined from the computed magnetic resonance maps. The antra of early atretic dominant follicles showed higher T(2)-weighted mean pixel value (P < 0.008) and heterogeneity (P < 0. 01) and lower T(2) heterogeneity (P < 0.008) than growing follicles. Subordinate follicles in the presence of a preovulatory dominant follicle had higher T(1), T(1) heterogeneity, proton density, proton density heterogeneity, and lower mean pixel value in T(1)-weighted images than subordinate follicles of the anovulatory wave (P < 0.04). T(1) relaxation rate heterogeneity and proton density heterogeneity were positively correlated with follicular fluid oestradiol concentration (r = 0.4 and 0.3; P < 0.04). T(2) relaxation rate heterogeneity was positively correlated with follicular fluid progesterone concentration (r = 0.4; P < 0.008). Quantitative differences in magnetic resonance image attributes of the antrum observed among phases of follicular development and regression coincided with changes in the ability of the dominant follicle to produce steroid hormones and ovulate, and thus were indicative of physiological status and follicular health.

The acquisition of unaliased ROSE (Radially Oriented Sinusoidal Excursions in k-space) data requires the appropriate selection of a gradient frequency, number of interleaves, and number of data samples per acquisition. When the data samples are uniformly distributed in time, they fall on irregularly spaced circles in k-space. Aliasing due to radial undersampling will be eliminated when the number of sample circles equals the number of pixels in one dimension of the desired reconstructed image. Azimuthal aliasing will be completely eliminated when the total number of ROSE petals is four times the number of pixels in one dimension, but acceptable reconstructions may be had with fewer petals. Magn Reson Med 44:129-136, 2000., (Copyright 2000 Wiley-Liss, Inc.)

Three-dimensional magnetic resonance imaging coupled with maximum intensity projection display, a technique usually reserved for magnetic resonance imaging angiography, is useful for the study of ovarian follicular growth. The ovaries of 19 cows were examined each day by transrectal ultrasonography. From these data, the precise phase of the ovarian cycle was determined and cows were ovariectomized on day 3 of wave one (n = 5), on day 6 of wave one (n = 4), on day 1 of wave two (n = 4), >/= 17 days after ovulation (n = 5), and on the day of ovulation (n = 1). The excised ovaries were examined by magnetic resonance imaging using a fast imaging with steady state precession imaging sequence with maximum intensity projection reconstruction, displayed as a cine-loop of the ovaries rotating in space. This provided the clearest view among the three principal three-dimensional steady state data acquisition approaches tried; the follicles and other ovarian structures could be distinguished unambiguously. Results from the bovine model indicate that the acuity of the three-dimensional fast imaging with steady state precession technique has potential application in in vivo intravaginal imaging in women for studying normal and pathological ovarian function.

The application of a knowledge-based segmentation method to the problem of automatically detecting the outer follicle wall boundary in ultrasonographic images of ovaries is presented. A combination of computer detection and interactive adjustment was used to define an approximate inner follicle-wall boundary, which was then used by the computer algorithm as a priori knowledge to automatically find the outer follicle-wall border. The segmentation algorithm was tested on ultrasonographic images of women's ovaries that were imaged in vivo. The semiautomatic segmentations were compared to segmentations by an expert human observer in terms of border placement differences and in terms of quantitative parameters relevant to the physiologic status of the follicles. These physiological parameters include total and specific signal intensity from the follicle and from the follicle wall. The computer-detected outer follicle wall boundaries correlated well with the human observer-defined wall boundaries, in terms of enclosed follicle area, specific and total follicle signal, enclosed wall area, and specific and total wall signal. The actual border placement differences were also small, with a maximum placement difference of 1.47 +/- 0.83 mm and a root mean square (r.m.s.) placement difference of 0.59 +/- 0.28 mm.