Combined MRI and MRA for limb salvage planning

Magnetic resonance angiography (MRA) was performed in two patients with primary bone neoplasms being staged for differing types of limb-salvage procedures. A gated two-dimensional (2D) phase contrast MRA sequence that is capable of variable velocity encoding throughout the cardiac cycle was used to acquire the MRA images. The resulting cardiac phase images were added with a matched filter algorithm to create a single high signal-to-noise ratio image. An SE MR image in the same plane and field of view was then added to the MRA image. This displayed the relationship of vascular and stationary tissues within a composite picture: a "2D combined angiographic-stationary tissue" image. The anatomy represented gave information equivalent to conventional angiography. While this technique is not an ideal 3D rendering, it is easier to use. Along with other SE sequences acquired as needed for staging, it provides a complete preoperative evaluation of a tumor bed or donor site for a vascularized graft harvest.

Quantitative velocity images from thick slab 2D phase contrast

Thick slab two-dimensional phase contrast (2D PC) angiography provides a large amount of anatomical information in a short acquisition time. Quantitative velocity information is, however, destroyed by the necessary projection dephasing gradient. We present a 2D PC acquisition scheme which retains quantitative velocity information in a thick slab acquisition. A thick slab acquisition produces an image which is a projection through an entire vessel. Accurate velocity measurements must take into account the intravoxel phase cancellation caused by a distribution of velocities within the vessel. In addition to the description of the data acquisition scheme, we outline a method for determining the mean velocity within a vessel having a laminar flow distribution. It is shown that without the consideration for intravoxel phase cancellation, the mean velocity measurement from thick slab images overestimates the true mean velocity by 10%.

Peripheral MR angiography with variable velocity encoding. Work in progress

An electrocardiographically triggered two-dimensional phase-contrast (PC) magnetic resonance angiographic pulse sequence was developed in which velocity encoding (VENC) was varied throughout an acquisition in response to changes in blood velocity during the cardiac cycle. This was done to better capture signal in the peripheral vasculature, where pulsatile flow degrades images. After reconstruction, a matched filter addition technique was applied to the cardiac phase images to obtain a single high-quality static image. Images were obtained of six healthy volunteers--with and without varying VENC--and contrast-to-noise ratio (C/N) calculations were performed for the added images. Varying VENC significantly improved vascular signal from small and large vessels (P less than .02), but it was most helpful for small vessels, for which the C/N increased by as much as 260% (average increase, 149%). These preliminary findings suggest that variable VENC can enhance the signal from the small and large peripheral blood vessels in cardiac-gated PC acquisitions.

A data adaptive reprojection technique for MR angiography

Inability to detect vessel overlap and vascular loops can compromise the interpretation of magnetic resonance angiograms. A data-adaptive ray tracing (DART) technique was developed to produce the appropriate variations in signal intensity at points of vessel overlap in order to simulate the standard angiographic representation of vessels. In this technique a threshold is utilized to identify vessels in the image slices composing a 3D angiographic data set. A mask, which defines regions slightly larger than the vessel boundaries, is obtained by blurring the vessel information surviving the initial threshold. This mask is converted to binary form prior to multiplication by the original angiographic data set. Following application of an additional threshold to the masked data, line integrals through the regions defined by the mask are performed to obtain an angiographic signal proportional to the integrated vessel signal as in conventional angiography. This integrated reprojection is then uniquely combined with a maximum intensity pixel (MIP) reprojection to produce the final DART image. The application of the DART technique to 2D time-of-flight and 3D phase-contrast angiograms successfully enabled the identification of over-lapping vessels and vascular loops. DART was also found to produce less vessel narrowing than the MIP technique.

Geometric quantitative coronary arteriography. A comparison of unsubtracted and dual energy-subtracted images

The application of dual energy (DE) subtraction techniques to quantitative coronary arteriography (QCA) has the advantage of removing the tissue signal surrounding the vessel profile. We have compared the performance of two geometric QCA algorithms on DE-subtracted and -unsubtracted images to determine, for each, if DE subtraction is advantageous. The two algorithms under study were an edge detection algorithm and a Fourier analysis-based algorithm. For each algorithm, linear regression analysis was performed of measured cross-sectional area (CSA) versus actual CSA of coronary vessel phantoms. The edge detection algorithm was found to have improved precision (P less than .05) when applied to the DE-subtracted images. The Fourier analysis algorithm, however, was not effected by the DE subtraction. Among the unsubtracted image results, the Fourier measurements were more accurate (P less than .05) than the edge detection measurements. We conclude that the benefits to edge detection QCA of DE tissue subtraction outweigh the disadvantages of increased image noise and possible misregistration artifacts. However, the Fourier algorithm is relatively insensitive to tissue signal variations.

Spontaneous lymphocyte proliferation in HTLV-I/II infection reflects preferential activation of CD8 and CD16/56 cell subsets

Previous studies have shown that lymphocytes from HTLV-infected persons spontaneously proliferate when cultured in vitro. We investigated which cell subsets become activated in this response. Mononuclear cells from 16 HTLV-seropositive former blood donors and 9 seronegative controls were cultured for 7 days; activation was then assessed by measuring DNA synthesis in cultured cells and by monitoring CD25 expression by CD3, CD4, CD8, CD19, and CD16/56 lymphocyte subsets. Of the 16 cultures of HTLV + donor cells, 10 showed spontaneous proliferation (Prol + group) and 6 did not (Prol - group). Cytofluorometric analysis revealed a significant increase in the fractions of CD8 cells and CD16/56 cells expressing CD25 for the Prol + group, compared to the Prol- and control groups. Similarly, the fractions of CD25 cells expressing CD8 or CD16/56 were significantly increased in the Prol + group. Although neither the fraction of CD4 cells expressing CD25 nor the fraction of CD25 cells expressing CD4 were increased for the Prol + group, the modal fluorescence intensity value for CD25 expression by CD4 cells was increased, suggesting some CD4 cell activation occurred as well. Blastoid cells were, on average, 79% CD25 +, whereas the sum of CD4 + CD25 + (27%), CD8 + CD25 + (30%), CD19 + CD25 + (3%), and CD16/56 + CD25 + (35%) subsets was 95%; the presence of 17% CD8 + CD16/56 + cells accounted for most of this discrepancy. These findings indicate that spontaneous lymphocyte proliferation in HTLV infection reflects preferential activation of CD8 and CD16/56 cell subsets, apparently including the minor CD8 + CD16/56 + subset.

Absolute diameter measurements of coronary arteries based on the first zero crossing of the Fourier spectrum

Current edge detection methods used to determine coronary artery dimensions from digital (DSA) images suffer a strong dependence on the system's modulation transfer function (MTF). The videodensitometric algorithms are less sensitive to MTF blurring, yet still result in an overestimation of the vessel size of 10% to 25% for blurring aperture sizes of 50% to 80% of the vessel diameter. We propose a new algorithm to measure the absolute diameter of a vessel which has a lower sensitivity to the system MTF for blurring aperture sizes up to 80% of the vessel diameter. A consequence of the similarity theorem of Fourier transform pairs is that the "width" of the Fourier transform, as characterized by the first zero crossing in frequency space, is inversely proportional to the width of the vessel profile. For an ideal (unblurred) vessel image, the width of the vessel profile is equal to the diameter of the vessel. For a blurred image this is not true. In frequency space, however, the transform of the blurred profile is simply the product of the transformed ideal profile and the system MTF. Thus, if the blurring aperture of the system is below some critical value, the first zero of the unblurred profile will still dominate the transform of the blurred profile. For vessels of circular cross section and a rectangular blurring aperture, this critical aperture size is approximately 80% of the vessel diameter. A more detailed explanation of the theory and calculations involved in this measurement, along with measurements of computer simulated and phantom vessels is presented.

Functional differences between small and large luteal cells of the late-pregnant vs. nonpregnant cow

This paper describes an in vitro model for the study of two types of steroidogenic luteal cells from cows in different physiological states. Two different populations of enzymatically dispersed bovine luteal cells were separated on the basis of size in a Cel-Sep Sedimentation Chamber. The separated small (12.5-23 micron in diameter) and large (greater than 23 micron in diameter) luteal cells of late-pregnant cows (Days 190-280) contained the distinct morphological characteristics previously defined for these two populations of cells. Cells were evaluated for progesterone (P4) production during a 3-h incubation with and without bovine luteinizing hormone (bLH, 10 ng/ml). Both small and large luteal cells from the late-pregnant cow were found to contain equal levels of P4 at Time 0 and increased but equal levels of P4 after a 3-h incubation. Neither cell type showed an increase in P4 production in response to the addition of bLH (p greater than 0.05). Since these results differed from earlier reports for luteal cells of the nonpregnant cow, small and large luteal cells of the mid-cycle (Day 14) were incubated, and the levels of P4 production were compared with P4 levels from the late pregnant cow. In agreement with previous reports for nonpregnant cows, progesterone content at Time 0 was 7-fold higher in large cells than in small cells (p less than 0.05), and after 3 h of incubation, 13-fold higher (p less than 0.05). Although the small cells responded to the presence of bLH in the incubation medium with a 4-fold increase in P4 production, this increase was not significant (p greater than 0.05). The large cell did not respond to bLH. However, the large cell type continued to contain and produce more P4 than did the small cells treated with bLH. This study indicates that both the small and large luteal cells of late-pregnancy are able to produce P4. However, the large luteal cell of the estrous cycle produces greater quantities of P4 than does the small luteal cell or the large luteal cell of late pregnancy.