Magnetic resonance imaging of soft-tissue tumors: Comparison with computed tomography

Twenty-seven patients with soft-tissue tumors were examined with a Picker 0.15-tesla resistive magnet and by computed tomography (CT). In all but one patient, MRI was better than or equal to CT in defining the anatomic extent of the tumor. We could determine whether major vascular structures were engulfed by the tumor in 80% of the MRI examinations but only in 62% of the CT scans. MRI and CT were equally effective in determining the presence or absence of bony invasion. The MRI images of all the tumors showed increased signal intensity relative to normal muscle when spin-echo (SE) sulse sequences with long repeat times were used (SE: echo time [TE], 60 ms; repetition time [TR], 2,000 ms). When T1 weighted pulse sequences were used (SE: TE, 30 ms; TR, 500 ms or inversion recovery: inversion time, 500 ms; TE, 40 ms; TR, 2,000 ms) the malignant tumors showed decreased signal intensity compared to normal muscle. Only lipomas showed high signal intensity on both T1 and T2 weighted pulse sequences.     

3D fast flair: a CSF-nulled 3D fast spin-echo pulse sequence

Since its introduction, the fluid attenuated inversion recovery (FLAIR) sequence has found many applications in the central nervous system (CNS), because of its heavy T₂ weighting and excellent cerebrospinal fluid (CSF) suppression. More recently fast spin-echo based variants have been developed that greatly reduce scan time; such sequences are often referred to as “fast FLAIR.” We present what we believe to be the first implementation of FLAIR using a three dimensional (3D) fast spin-echo-based pulse sequence, which combines the cerebrospinal fluid suppression and good lesion contrast of two-dimensional fast FLAIR with the advantages of a three-dimensional sequence such as higher signal-to-noise ratio (SNR) per unit time, thinner slices (giving reduced partial volume effect) and the ability to reformat the data in an arbitrary plane.     

Elimination of chemical shift artifacts of thoracic spine with contrast-enhanced FLAIR imaging with fat suppression at 3.0 T

The purpose of this study was to assess the effect of chemical shift artifacts and fat suppression between contrast-enhanced T1-weighted fast spin-echo (FSE) sequence with fat suppression and contrast-enhanced T1-weighted fluid attenuated inversion recovery (FLAIR) sequence with fat suppression in magnetic resonance imaging (MRI) of the thoracic spine at 3.0T. Forty patients, who underwent MRI examination, were recruited and analyzed both qualitatively and quantitatively. Due to chemical shift artifacts in the T1-weighted FSE, 14 of the patients were found to be of non-diagnostic value. On the contrary, in 11 of those 14 patients, no chemical shift artifacts were observed in the T1-weighted FLAIR sequence. Regarding the efficiency of fat suppression, both sequences achieved successful fat suppression. Consequently, the use of T1-weighted FLAIR fat suppression after contrast administration sequence seems to eliminate or significantly reduce image quality deterioration stemming from chemical shift artifacts in thoracic spine examinations.     

Magnetization transfer contrast (MTC) and long repetition time spin-echo MR imaging in multiple sclerosis

Purpose: To study whether application of magnetization transfer contrast (MTC) improves visibility and detection of multiple sclerosis (MS) lesions on long repetition time (TR) conventional spin-echo (CSE) or fast spin-echo (FSE) magnetic resonance (MR) imaging.Material and methods: In 20 patients and 5 controls, MR images were obtained using long repetition time CSE and FSE sequences with and without MTC. Signal-to-noise ratios of normal appearing white matter (NAWM) and selected lesions, and contrast-to-noise ratios between lesions and NAWM, were calculated. Lesions were counted and total lesion volume was measured in a blinded fashion for each sequence.Results: In controls, MT effect in white matter (16.3% vs. 12.2%) was higher for CSE than for FSE (p < 0.01). Application of MTC to either CSE or FSE resulted in a significantly lower decrease in signal intensity of NAWM in patients compared to white matter in controls (p < 0.01). Furthermore, in patients signal intensity of lesions was less decreased than signal intensity of NAWM (p < 0.01). Compared to sequences without MTC, contrast-to-noise ratios were significantly higher on both CSE (10.9%) and FSE (6.3%) when MTC was applied (p < 0.01). Despite better visibility, the number of lesions detected on either sequences did not increase when MTC was applied. For CSE with MTC, we found an almost equal number of lesions and for FSE with MTC, we found even less lesions (p < 0.01). Total lesion volume did not change significantly when MTC was applied.Conclusion: Although contrast between lesions and NAWM improved when magnetization transfer contrast was applied, this did not increase detection of MS lesions on either CSE or FSE MR imaging.     

Bone marrow imaging using STIR at 0.5 and 1.5 T.

We retrospectively examined MR images in 82 patients to evaluate the usefulness of short inversion time inversion recovery (STIR) in bone marrow imaging at 0.5 and 1.5 T. The study included 56 patients at 1.5 T and 26 patients at 0.5 T with a variety of pathologic bone marrow lesions (principally oncological), and compared the contrast and image quality of STIR imaging with spin-echo short repetition time/echo time (TR/TE), long TR/TE, and gradient-echo sequences. The pulse sequences were adjusted for optimal image quality, contrast, and fat nulling. STIR appears especially useful for the evaluation of red marrow (e.g., spine), where contrast between normal and infiltrated marrow is greater than with either gradient-echo or T1-weighted images. STIR is also extremely sensitive for evaluation of osteomyelitis, including soft tissue extent. In more peripheral (yellow) marrow, T1-weighted images are usually as sensitive as STIR. Limitations of STIR include artifacts, in particular motion artifact that at high field strength necessitates motion compensation. At 0.5 T, however, motion compensation is usually not necessary. Also, because of extreme sensitivity to water content, STIR may overstate the margins of a marrow lesion. With these limitations in mind, STIR is a very effective pulse sequence at both 0.5 and 1.5 T for evaluation of marrow abnormalities.     

MRI with superparamagnetic iron oxide: efficacy in the detection and characterization of focal hepatic lesions

The purpose of this study was to evaluate the potential of superparamagnetic iron oxide particles (SPIO) as tissue specific contrast agent in magnetic resonance (MR) imaging in detection and characterization of focal hepatic lesions. We investigated 45 patients with focal hepatic lesions. T1-weighted SE (TR 650/TE 15 ms) and T2-weighted SE (TR 2015-2030/TE 45 and 90 ms) unenhanced images were obtained. After SPIO application we performed T1-weighted images with and T2-weighted images with and without fat suppression using the same image parameters. Liver signal intensity decreased by 74% (min 47%, max 83%) on T2-weighted images after application of the contrast agent. Benign lesions (FNH, adenoma) showed an average signal drop of 40% (min 20%, max 47%) whereas malignant lesions showed no significant change of signal intensity on post-contrast images. The mean tumor-to-liver contrast-to-noise ratio (C/N) was improved in all post-contrast sequences irrespective of the lesion type. An additional increase of tumor-to-liver contrast by use of fat suppression technique could be established in the slightly T2-weighted sequence (TE 45 ms). In metastases, divided in different size groups, we could determine a significant size relation of tumor-to-liver C/N. After SPIO application the number of detected lesions increased distinctly, especially small foci are more easily demonstrated. SPIO particles are a efficacious contrast agent for MR examinations of the liver. For tumor characterization T1- and T2-weighted pre- and post-contrast images are necessary. The T1-weighted sequences are helpful to differentiate benign lesions such as cysts and hemangiomas from malignant lesions. Detection and differential diagnoses of hepatic lesions are improved by use of the SPIO-particles.     

A prospective assessment of breath-hold fast spin echo and inversion recovery fast spin echo techniques for detection and characterization of focal hepatic lesions

The purpose of this study was to prospectively assess two breath-hold T(2)-weighted fast spin-echo sequences and two breath-hold inversion recovery fast spin-echo sequences to determine their relative ability to detect and characterize focal hepatic lesions. Fourteen patients with a total of nineteen proven focal hepatic lesions were imaged with two breath-hold T(2)-weighted (T2W) fast spin echo sequences (HASTE TE = 66 and HASTE TE = 120), two breath-hold inversion recovery fast spin echo sequences (IRFSE TE = 64 and IRFSE TE = 95), and a nonbreath-hold T(2)-weighted fast-spin echo sequence (FSE TE = 96-120). Contrast-to-noise ratios (CNRs) were measured for all proven lesions on all sequences. Both IRFSE sequences and the HASTE sequence with TE = 66 showed an improvement in lesion-liver and liver-spleen CNRs compared to the nonbreath-hold T2W sequence. The mean difference in CNR between benign and malignant lesions was largest for the HASTE TE = 120 sequence. These preliminary results suggest that a breath-hold IRFSE sequence (TE = 64 or 95) has an equal ability to detect focal hepatic lesions as a nonbreath-hold T2W FSE sequence (TE = 96-120). The HASTE TE = 120 showed the greatest ability to discriminate between benign and malignant lesions.     

Fast flair imaging of the brain using the fast spin-echo and gradient spin-echo technique

The purpose of this study was to compare the gradient spin-echo (GRASE) to the fast spin-echo (FSE) implementation of fast fluid-attenuated inversion recovery (FLAIR) sequences for brain imaging. Thirty patients with high signal intensity lesions on T₂-weighted images were examined on a 1.5 T MR system. Scan time-minimized thin-section FLAIR-FSE and FLAIR-GRASE sequences were obtained and compared side by side. Image assessment criteria were lesion conspicuity, contrast between different types of normal tissue, image quality, and artifacts. In addition, contrast ratios and contrast-to-noise ratios were determined. Compared to FSE, the GRASE technique allowed a 17% reduction in scan time but conspicuity of small lesions in particular was significantly lower on FLAIR-GRASE images because of higher image noise and increased artifacts. Gray-white differentiation was slightly worse on FLAIR-GRASE. Physiological ferritin deposition appeared slightly darker on FLAIR-GRASE images and susceptibility artifacts were stronger. Fatty tissue was less bright with FLAIR-GRASE. With current standard hardware equipment, the GRASE technique is not an adequate alternative to FSE for the implementation of fast FLAIR sequences in routine clinical MR brain imaging.     

Brain imaging: Comparison of T1W FLAIR BLADE with conventional T1W SE

IntroductionAlthough T1 weighted spin echo (T1W SE) images are widely used to study anatomical details and pathologic abnormalities of the brain, its role in delineation of lesions and reduction of artifacts has not been thoroughly investigated. BLADE is a fairly new technique that has been reported to reduce motion artifacts and improve image quality.ObjectiveThe primary objective of this study is to compare the quality of T1-weighted fluid attenuated inversion recovery (FLAIR) images with BLADE technique (T1W FLAIR BLADE) and the quality of T1W SE images in the MR imaging of the brain. The goal is to highlight the advantages of the two sequences as well as which one can better reduce flow and motion artifacts so that the imaging of the lesions will not be impaired.Materials and methodsBrain examinations with T1W FLAIR BLADE and T1W SE sequences were performed on 48 patients using a 1.5 T scanner. These techniques were evaluated by two radiologists based on: a) a qualitative analysis i.e. overall image quality, presence of artifacts, CSF nulling; and b) a quantitative analysis of signal-to-noise ratios (SNR), contrast-to-noise ratios (CNR) and Relative Contrast. The statistical analysis was performed using the Kruskal-Wallis non-parametric system.ResultsIn the qualitative analysis, BLADE sequences had a higher scoring than the conventional sequences in all the cases. The overall image quality was better on T1W FLAIR BLADE. Motion and flow-related artifacts were lower in T1W FLAIR BLADE. Regarding the SNR measurements, T1W SE appeared to have higher values in the majority of cases, whilst T1W-FLAIR BLADE had higher values in the CNR and Relative Contrast measurements.ConclusionT1W FLAIR BLADE sequence appears to be superior to T1W SE in overall image quality and reduction of motion and flow-pulsation artifacts as well as in nulling CSF and has been preferred by the clinicians. T1W FLAIR BLADE may be an alternative approach in brain MRI imaging.     

Sequence parameters of double spin-echo sequences affect quantification of citrate

Quantification of citrate by localized ¹H spectroscopy is usually performed using the water signal as reference, but the signal behavior of the J-coupled AB spin system of citrate after multipulse excitation is not as trivial as for uncoupled substances. The influence of the timing scheme of double spin-echo sequences and of the spatial flip angle distribution of (nonideal) refocusing pulses was analyzed systematically for the citrate resonances. Both single echo times of the double spin-echo sequence were varied between 20 ms and 250 ms in theoretical and experimental approaches. Relatively long total echo times (TE > 120 ms) provide high selectivity to citrate signals, since signals from triglycerides at 2.6 ppm are markedly reduced. Asymmetrical timing schemes of the double spin-echo sequence with one short single echo time of 20 ms and one longer single echo time of about 120 ms result in high integral signal from the central lines of citrate, whereas symmetrical timing leads to high sensitivity for total echo times TE near 100 ms. The integral citrate signals in spectra with relatively long echo times (TE > 120 ms) were found to depend markedly on the type of the refocusing pulses, affecting quantitative citrate measurements in vitro and in vivo.     

Feasibility of fast MR imaging of the liver at 1.5 T

A phantom with T1 and T2 relaxation times encompassing normal liver and liver lesions was constructed to evaluate fast magnetic resonance pulse sequences using TR from 21-100 milliseconds, TE 12-60 milliseconds and flip angles from 5 degrees-90 degrees. Ten of these fast MR sequences were then selected and compared with conventional spin-echo sequences in normal volunteers (n = 3) and in patients with liver lesions (n = 6). Subjectively, the fast MR sequences eliminated motion artefacts. Objectively, 8 of 10 fast sequences had signal-to-noise ratios comparable to spin-echo imaging whereas only 2 of 10 had contrast-to-noise ratios that were similar to spin-echo imaging. This preliminary study, performed at 1.5 Tesla, does not show any clear-cut advantage of fast imaging over spin-echo imaging in the detection of liver lesions. The use of a liver tissue equivalent phantom provides a rapid, practical approach in evaluation of fast scans.     

Dual-echo breathhold T(2)-weighted fast spin echo MR imaging of liver lesions

The purpose of this study was to develop a multi-shot dual-echo breathhold fast spin echo technique (DFSE) and compare it with conventional spin echo (T2SE) for T(2)-weighted MR imaging of liver lesions. The DFSE acquisition (EffTE1/EffTE2/TR = 66/143/2100 ms) imaged 5 sections per 17 s breathhold. T2SE imaging (TE1/TE2/TR = 60/120/2500 ms) required 16:55 (min:s) for 14 sections. Both techniques used a receive-only phased-array abdominal multicoil and provided 192 x 256 effective resolution. The results showed first and second echo relative DFSE/T2SE contrast values for 27 representative lesions (15 consecutive patients) were 1.08 +/- 0.05 and 1.16 +/- 0.09 (mean +/- STD mean), respectively. Corresponding CNR values were 1.12 +/- 0.09 and 0.97 +/- 0.12. Overall DFSE was comparable-to-superior to T2SE for lesion sizing and image artifact. DFSE lesion detection was inferior to T2SE's in several patient studies because of decreased conspicuity of lesions located near multicoil edges and because of poor breathhold-to-breathhold reproducibility and lack of breathholding. However both DFSE (and T2SE) provided lesion detection rated to be of diagnostic quality for all patient studies. In conclusion, we found that DFSE provides diagnostically useful dual-echo T(2)-weighted MR liver images in a greatly decreased acquisition time.     

Signal characteristics of FLAIR related to water content: comparison with conventional spin echo imaging in infarcted rat brain

To examine the correlation between tissue water content and signal intensity on fluid-attenuated inversion recovery (FLAIR) images, we analyzed infarcted rat brain, verified the results by theoretical simulation, and compared them with conventional spin-echo images. We produced brain infarction with cavitation in five rats by middle cerebral artery occlusion. After in vivo MRI, histologic sections of the MRI plane were obtained. We measured the signal intensity of regions on FLAIR and spin-echo images, and measured the area of cavitation on histologic sections. We plotted curves of cavity percentage to signal intensity. Theoretical values were calculated using a two-compartment model. On the curve of cavity area to signal intensity, the signal on FLAIR images peaked in tissues with 20% to 30% area of cavitation. On the theoretical curve, the signal on FLAIR images peaked at 90% tissue water content. These results seem to be characteristic of FLAIR.     

Evaluation of the prostate by magnetic resonance imaging

Forty-seven male patients with suspected prostatic disease underwent magnetic resonance imaging (MRI) of the pelvis on a Picker resistive magnet operating at 0.15 T; 33 had histologically proved adenocarcinoma, 12 benign prostatic hypertrophy, 1 a transitional cell carcinoma, and 1 a seminoma. Eleven normal subjects also were included in the study. The study attempted to (1) define the MRI characteristics of the normal prostate, benign prostatic hypertrophy, and prostatic adenocarcinoma, (2) evaluate various pulse sequences in imaging the prostate, and (3) compare MRI findings with clinical, pathologic, and computed tomography results. Various pulse sequences, including inversion recovery and spin-echo with short and long TE and TR, were used. MRI was sensitive in detecting intracapsular and extracapsular prostatic disease. The finding of inhomogeneous signal texture throughout the gland was a sensitive but nonspecific finding for adenocarcinoma. A focal nodule with prolonged T1 and T2 relaxation times was the most specific MRI finding for adenocarcinoma. Extracapsular spread of neoplasm was often demonstrated, and because of its superior soft-tissue contrast ability, MRI was more accurate than computed tomography in delineating extracapsular extension.     

Application of multiple inversion recovery for suppression of macromolecule resonances in short echo time (1)H NMR spectroscopy of human brain.

Macromolecules contribute broad "background" resonances to the (1)H NMR brain spectra at short echo times. The application of long echo times is the most widely used method for removing these resonances. Here, it is demonstrated that these background resonances may be suppressed at short echo times using multiple inversion recovery (MIR). In the technique presented, the MIR sequence consists of four adiabatic inversion pulses, applied preparatory to a 20-ms echo time stimulated echo localization sequence. The inversion times (359, 157, 69, and 20 ms) were selected to preferentially suppress macromolecules with longitudinal relaxation times between 38 and 300 ms. While the resulting spectra have lower overall signal-to-noise, baseline contributions from macromolecules are greatly reduced. Unlike the typical long TE acquisitions, the short TE MIR acquisition preserves the myo-inositol resonance.     

Relaxation times and NMR signals

The strength of signals in magnetic resonance imaging (and the resulting image contrast) depends not just on the number density of the nuclei being detected, but also on the relaxation times, T1 and T2. The relationship of signal strength to relaxation time depends on the particular choice of pulse sequences used to produce the signals. The effects of the T1 relaxation time on signal strength are discussed for the commonly used imaging techniques "partial saturation" and "inversion recovery." Production of spin echos and the effect of the T2 relaxation time on spin-echo signal strength are also discussed.     

Improved contrast at 1.5 tesla using half-Fourier imaging: application to spin-echo and angiographic imaging

Half-Fourier imaging is useful for reducing imaging time by requiring less than the usual number of phase-encoding steps. This increase in speed can be traded off for longer repeat times, TR, for improved contrast-to-noise in the same imaging time or to collect short asymmetric echoes. Consequently, it is shown to be especially useful for long TR spin-echo imaging where at 1.5 T a repeat time of 4 sec is recommended for a double-echo TE = 30/90 sequence or 3 sec for a double-echo TE = 15/90 sequence. Short TR FLASH imaging also benefits from a longer TR since there is more time to spoil the signal. In both cases, there is the advantage when a multislice acquisition mode is used that more slices (and hence, a larger volume) can be taken. Another application is to apply half-Fourier imaging in the read direction to avoid spin dephasing and motion artifacts. This is particularly useful in angiographic imaging where smaller pixel sizes and shorter echo times both reduce pixel dephasing. Again, even though taking less than the usual number of data points leads to a reduction in S/N, the improved signal and resolution for blood vessels can more than compensate this loss.     

MRI-monitored cryosurgery in the rabbit brain

The inability to observe the transient, irregular shape of the frozen region that develops during cryosurgery has inhibited the application of this surgical technique to the treatment of tumors in the brain and deep in visceral organs. We used proton NMR spin-echo and spoiled gradient-echo imaging to monitor the development of frozen lesions during cryosurgery in the rabbit brain and the resulting postervosurgical changes up to 4 hr after freezing. Spoiled gradient-echo images (TE = 14 ms; TR = 50 ms) were acquired during freezing and thawing at a rate of 15 s/slice. Although the frozen region itself is invisible in MR images, its presence is distinguished easily from the surrounding unfrozen soft tissue because of the large contrast difference between frozen and unfrozen regions. T₂-weighted spin-echo images (TE = 100 ms, TR = 2 s) obtained after thawing suggest that edema forms first at the margin of the region that was frozen (cryolesion) and then moves into the region's core. Histological examination showed complete necrosis in the cryolesion and a sharp transition to undamaged tissue at the margin of the lesion and its image. Blood-brain barrier (BBB) damage was investigated using gadolinium-DTPA. The region of edema in the T₂-weighted spin-echo images was coincident with the area of BBB damage in the Gd-DTPA-enhanced T₁-weighted spin-echo images (TE = 33 ms, TR = 400 ms) and both were distinguishable as areas of high signal relative to the surrounding normal tissue. The results of these experiments indicate that MR can both effectively monitor the cryosurgical freezing and thawing cycle and characterize the postcryosurgical changes in tissue during follow-up.     

Disappearance of tumor contrast on contrast-enhanced FLAIR imaging of cerebral gliomas

Contrast-enhanced fluid-attentuated inversion recovery (FLAIR) magnetic resonance (MR) imaging has shown to be a valuable diagnostic modality in the assessment of cerebral gliomas. In this study we report of a potential pitfall regarding the delineation of enhancing tumor parts on contrast enhanced FLAIR imaging. In a limited number of patients, the administration of gadolinium obscures the area of contrast enhancement on contrast enhanced FLAIR images. Therefore the delineation of the macroscopic tumor parts, which are of great importance for the treatment planning is substantially worsened.     

Are multi-contrast magnetic resonance images necessary for segmenting multiple sclerosis brains? A large cohort study based on deep learning

BackgroundMagnetic resonance images with multiple contrasts or sequences are commonly used for segmenting brain tissues, including lesions, in multiple sclerosis (MS). However, acquisition of images with multiple contrasts increases the scan time and complexity of the analysis, possibly introducing factors that could compromise segmentation quality.ObjectiveTo investigate the effect of various combinations of multi-contrast images as input on the segmented volumes of gray (GM) and white matter (WM), cerebrospinal fluid (CSF), and lesions using a deep neural network.MethodsU-net, a fully convolutional neural network was used to automatically segment GM, WM, CSF, and lesions in 1000 MS patients. The input to the network consisted of 15 combinations of FLAIR, T1-, T2-, and proton density-weighted images. The Dice similarity coefficient (DSC) was evaluated to assess the segmentation performance. For lesions, true positive rate (TPR) and false positive rate (FPR) were also evaluated. In addition, the effect of lesion size on lesion segmentation was investigated.ResultsHighest DSC was observed for all the tissue volumes, including lesions, when the input was combination of all four image contrasts. All other input combinations that included FLAIR also provided high DSC for all tissue classes. However, the quality of lesion segmentation showed strong dependence on the input images. The DSC and TPR values for inputs with the four contrast combination and FLAIR alone were very similar, but FLAIR showed a moderately higher FPR for lesion size <100 μl. For lesions smaller than 20 μl all image combinations resulted in poor performance. The segmentation quality improved with lesion size.ConclusionsBest performance for segmented tissue volumes was obtained with all four image contrasts as the input, and comparable performance was attainable with FLAIR only as the input, albeit with a moderate increase in FPR for small lesions. This implies that acquisition of only FLAIR images provides satisfactory tissue segmentation. Lesion segmentation was poor for very small lesions and improved rapidly with lesion size.