R. M. Henkelman

Characterizing white matter with magnetization transfer and T2

A magnetization‐transfer (MT) CPMG hybrid experiment was performed to analyze T2 relaxation and MT characteristics in bovine optic nerve. Two exchanging liquid pools with their own, independent MT characteristics were necessary to model both the T2 relaxation and the MT data. The model agrees well with the experimental data and yields physically realistic parameters. The MT effect for myelin water is approximately nine time larger than that for intra/intercellular water, indicating that the MT characteristics observed for white matter are mainly related to myelin. The model can be used to probe parameters that would be difficult to achieve experimentally. The exchange process between the two tissue compartments does not drastically affect the amplitudes and relaxation rates of the T2 components, but is fast enough to significantly influence their MT characteristics. Although, both the MT and T2 experiments described in this paper are too time consuming to be applied in routine clinical work, presented results can be useful in interpreting clinical pulse sequences that are sensitive to myelin. Magn Reson Med 42:1128–1136, 1999.

Noise and filtration in magnetic resonance imaging

Noise in two-dimensional Fourier transform magnetic resonance images has been investigated using noise power spectra and measurements of standard deviation. The measured effects of averaging, spatial filtering, temporal filtering, and sampling have been compared with theoretical calculations. The noise of unfiltered images is found to be white, as expected, and the choice of the temporal filter and sampling interval affects the noise in a manner predicted by sampling theory. The shapes of the imagers spatial frequency filters are extracted using noise power spectra.

Quantifying tissue damage due to focused ultrasound heating observed by MRI

Focused ultrasound heating of ex vivo bovine kidney and liver was monitored using magnetic resonance imaging (MRI) to investigate the quantitative relationship between time‐dependent temperature elevations and altered contrast in MR images due to thermal coagulation. Proton resonance frequency shift MR thermometry was performed during heating at 10 sec intervals (single‐slice fast spoiled GRASS [FSPGR], θ/TE/TR 30o/11/39 msec, field of view 8 cm, 256 × 256, 3 mm slice thickness, 1 NEX); post‐heating MR images were T1‐weighted (3D‐FSPGR, θ/TE/TR 60o/25/200 msec, 1 mm slice thickness, 3 NEX). Analysis of the resulting temperature versus time data using the Arrhenius relationship and a simple binary discrimination model showed that thermal coagulation occurred with heating at approximately 54°C for 10 sec in both tissues and could be predicted with approximately 625 μm spatial resolution. These results suggest that quantitative MR guidance of thermal coagulation therapy is feasible, and they provide information useful for designing future investigations in vivo.Magn Reson Med 41:321–328, 1999.

Analysis of changes in MR properties of tissues after heat treatment

To characterize changes in the MR parameters of tissues due to thermal coagulation, a series of T1, T2, diffusion, and magnetization transfer measurements were performed on a variety of ex vivo tissues: murine slow twitch skeletal muscle, murine cardiac muscle, murine cerebral hemisphere, bovine white matter, murine liver tissue, bovine retroperitoneal adipose tissue, hen egg white, human prostate and human blood. Standardized heat treatments were performed for each tissue type, over the temperature range from 37°C to 90°C. For all tissues, changes in each MR measurement resulting from thermal coagulation were observed above a threshold temperature of approximately 60°C. These changes are explained based on biophysical knowledge of thermal damage mechanisms and the MR properties of normal tissues, and are particularly relevant for interpreting the changes in image contrast that are observed when MRI is used to guide and monitor thermal coagulation therapy procedures. Magn Reson Med 42:1061–1071, 1999.

Comparison of thermal damage calculated using magnetic resonance thermometry, with magnetic resonance imaging post-treatment and histology, after interstitial microwave thermal therapy of rabbit brain

Clinical application of high-temperature thermal therapy as a treatment for solid tumours requires an accurate and close to real-time method for assessing tissue damage. Imaging methods that detect structural changes during heating may underestimate the extent of thermal damage. This is due to the occurrence of delayed damage manifested at tissue locations exposed to temperatures lower than those required to cause immediate structural changes. An alternative approach is to measure temperature and then calculate the expected damage based on the temperature history at each tissue location. Magnetic resonance (MR) imaging methods now allow temperature maps of the target and surrounding tissues to be generated in almost real-time. The aim of this work was to evaluate whether thermal damage zones calculated on the basis of MR thermometry maps measured during heating correspond to actual tissue damage as measured after treatment by histological methods and MR imaging. Four male rabbits were treated with high-temperature thermal therapy delivered in the brain by a single microwave antenna operating at 915 MHz. MR scanning was performed before, during and after treatment in a 1.5 T whole-body scanner. Temperature maps were produced using the proton resonance frequency (PRF) shift method of MR thermometry. In addition, conventional T1-weighted and T2-weighted spin-echo images were acquired after treatment. Thermal damage zones corresponding to cell death, microvascular blood flow stasis and protein coagulation were calculated using an Arrhenius analysis of the MR temperature/time course data. The calculated zones were compared with the lesions seen on histopathological examination of the brains which were removed within 6-8 h of treatment. The results showed that calculated damage zones based on MR thermometry agreed well with areas of damage as assessed using histology after heating was completed. The data suggest that real-time calculations of final expected thermal damage based on an Arrhenius analysis of MR temperature data may provide a useful method of real-time monitoring of thermal therapy when combined with conventional T2-weighted images taken after treatment.

Magnetic resonance imaging of temperature changes during interstitial microwave heating: A phantom study

Changes in magnetic resonance (MR) signals during interstitial microwave heating are reported, and correlated with simultaneously acquired temperature readings from three fiber-optic probes implanted in a polyacrylamide gel phantom. The heating by a MR-compatible microwave antenna did not interfere with simultaneous MR image data acquisition. MR phase-difference images were obtained using a fast two-dimensional-gradient echo sequence. From these images the temperature-sensitive resonant frequency of the 1H nuclei was found to decrease approximately by 0.008 ppm/ degree C. The method and results presented here demonstrate that noninvasive MR-temperature imaging can be performed simultaneously with interstitial microwave thermal treatment.

Proton relaxation enhancement by manganese(III)TPPS4 in a model tumor system

Managanese(III)tetraphenylporphine sulfonate [Mn(III)TPPS4] has been investigated as a tumor specific paramagnetic contrast agent for magnetic resonance imaging (MRI) of L1210 solid tumors in mice. Mn(III)TPPS4 was found to clear rapidly from the blood and concentrate in the kidneys, tumor and liver. Although relatively high ratios of tumor to normal tissues could be obtained (e.g., greater than 90 for tumor/muscle), the kidneys were found to have the highest concentration of the metalloporphyrin at all doses and time periods tested. A significant decrease in the longitudinal relaxation time was measured for excised tissues (kidney, tumor, liver, muscle) from mice that were treated with Mn(III)TPPS4. A linear correlation was observed between the longitudinal relaxation rate determined for L1210 tumor and the corresponding concentration of Mn(III)TPPS4 found at various injected doses and time intervals between the injection and analysis. A small animal radiofrequency receiver coil designed for use with a 0.15-T clinical imager was employed to evaluate the ability of Mn(III)TPPS4 to selectively increase the signal intensity of the implanted L1210 tumor. The images show a conspicuous enhancement in the contrast between the tumor and adjacent tissue upon treatment with this agent. The results indicate that Mn(III)TPPS4 is a useful prototype paramagnetic metalloporphyrin MRI contrast agent with a significant affinity for the L1210 tumor.

Application of autoregressive modelling in magnetic resonance imaging to remove noise and truncation artifacts

Magnetic resonance imaging data is conventionally reconstructed using two dimensional discrete Fourier transforms. However, there is growing interest in other types of spectral estimation which minimize noise and artifacts due to truncated data. This note presents preliminary results--showing the improvement obtainable using a modified autoregressive model, the Transient Error method.

Can MTR be used to assess cartilage in the presence of Gd-DTPA2–?

Magnetization transfer (MT) and T1 and T2 relaxation of normal, trypsinized, and interleukin‐1β (IL‐1β)‐treated cartilage were measured in the absence and presence of Gd‐DTPA2–. Without the addition of Gd‐DTPA2–, neither T1 nor T2 showed any significant change with cartilage damage. However, with Gd‐DTPA2–, trypsinized cartilage exhibited substantially shorter T1 than normal cartilage, as expected due to the glycosaminoglycan (GAG) loss in these samples, and associated increased Gd‐DTPA2– concentration. The T2 results were similar, but less dramatic. The MT pseudo first‐order exchange rate, RM0B, did not depend on the contrast agent concentration, as expected, and was significantly faster for trypsinized and slower for IL‐1β‐treated cartilage. In both cases, the MT fraction of the macromolecular pool M0B decreased while only trypsinized cartilage showed an increase in MT exchange rate R. The MT ratio (MTR) decreased with increasing Gd‐DTPA2– concentration. However, interpretation of the MTR results in the presence of Gd‐DTPA2– was complicated due to competing effects of increased longitudinal relaxivity and MT exchange. Therefore, in a cartilage sample with an unknown degree of GAG depletion and some collagen damage, a full MT analysis might be used to probe the molecular state of cartilage, but it would not be possible to use a simple MTR measurement after the administration of Gd‐DTPA2– to differentially determine the amount of cartilage degradation in the sample. Magn Reson Med 48:1081–1084, 2002.

Very slow in-plane flow with gradient echo imaging

It is well appreciated that gradient-echo imaging techniques with short recovery times are highly sensitive to flow. This article analyzes the effect of in-plane flow in gradient recalled acquisition in the steady state (GRASS). It is shown that there is loss of signal due to velocity-dependent dephasing effects at velo cities as slow as 0.2 mm per second. It is also shown that striations appear in GRASS images of flow phantoms. This effect, which has not been previously described, arises from a modulation of K-space in the phase-encoding direction during the transient approach to steady state. Although these bands can give the appearance of flow lines, they are completely artifactual and not readily interpretable. Thus, the appearance of in-plane fluid movement in clinical GRASS images is a complex combination of signal loss due to dephasing and artifactual banding. Therefore, the interpretation of flow in GRASS images should be attempted only with caution.