Cerebral Venous Thrombosis: MR Black-Blood Thrombus Imaging with Enhanced Blood Signal Suppression

Study Population

Twenty-six participants, including 13 healthy volunteers (9 women; 35–65 years of age; mean age, 53 years) and 13 patients (7 women; 19–54 years of age; mean age, 35 years), were recruited to undergo MR BTI studies. The inclusion criteria were no history of cerebrovascular diseases for healthy controls and CVT diagnosed on the basis of clinical symptoms and conventional imaging techniques (CT, TSE, SWI, contrast-enhanced [CE]-MRV, and TOF-MRV) for patients. Exclusion criteria included contraindications for MR imaging and intolerance to additonal MR images. Xuanwu Hospital institutional review committee approval and patients' informed consent were obtained.

Imaging System

Imaging was performed on a 3T whole-body system (Magnetom Verio; Siemens, Erlangen, Germany). A standard head-neck 12-channel coil was used for receiving signals from a volume ranging from the superior sagittal sinus to the internal jugular veins.

MR Imaging Sequence

The sequence used for MR BTI was designed on the basis of sampling perfection with application-optimized contrasts by using different flip angle evolution (SPACE sequence; Siemens).⁷ A previously proposed black-blood preparation method, delay alternating with nutation for tailored excitation (DANTE),⁹ was used to improve the suppression of venous flow signals while introducing minimal T2 weighting.^10,11 The DANTE module consists of a train of short, hard radiofrequency pulses interspersed with dephasing gradients that are applied simultaneously in all 3 orthogonal directions. The module is followed by chemically selective fat saturation and a T1-weighted SPACE readout. Whole-head spatial coverage is achieved using nonselective hard radiofrequency pulses for excitation, which averts the need for multiple signal averages to suppress free-induction-decay artifacts.⁷ A saggitally oriented imaging volume is prescribed, which requires fewer partitions than other imaging orientations, to further reduce the imaging time. Three 40-mm-wide spatial presaturation bands are applied immediately before excitation to suppress the signals from the nose and ears if they are located outside the prescribed imaging volume. The parameters of the DANTE preparation were optimized for adequate flow signal suppression in a separate volunteer study (detailed in the On-line Appendix).

Imaging Protocol

All participants underwent conventional MR imaging sequences followed by 2 randomized BTI scans, with DANTE preparation switched on (DANTE+) and off (DANTE−), respectively. The major imaging parameters used in the SPACE readout were the following: TR/TE = 600/14 ms; FOV = 300 × 206 × 162 mm³; matrix size = 384 × 264 × 208; spatial resolution = 0.78 mm isotropic; 6/8 partial Fourier in the partition-encoding direction; echo-train length = 36; parallel imaging (generalized autocalibrating partially parallel acquisition) acceleration factor = 2 in the phase-encoding direction; elliptic k-space sampling; scan time = 5 minutes 40 seconds.

Image Analysis

Image review and signal intensity measurements were performed on a workstation (syngo MultiModality Workplace; Siemens), where multiplanar reformation, maximum intensity projection, and minimum intensity projection functionalities were available to image reviewers. The venous system was divided into the following 16 segments: superior sagittal sinus, inferior sagittal sinus, straight sinus, confluence of sinuses, right transverse sinus, left transverse sinus, right sigmoid sinus, left sigmoid sinus, vein of Galen, internal cerebral vein, basal vein of Rosenthal, vein of Labbé, right cortical vein, left cortical vein, right internal jugular vein, and left internal jugular vein.

For the healthy volunteer group, both BTI DANTE+ and DANTE− image sets of each subject were reviewed side by side by a radiologist (G.W.) with 8 years of experience in MR image interpretation to identify residual blood signals in individual segments. When residual blood was observed in a segment, mean signal intensity was measured. The mean signal intensity of the adjacent brain parenchyma and normal lumen and noise (σ, SD of signal intensity in the adjacent air space) were also measured. The contrast-to-noise ratio (CNR) (A-to-B CNR = [SI_A–SI_B] /σ), where SI indicates signal intensity, was calculated for residual blood versus normal lumen (indicating the blood-suppressing performance of DANTE) and for brain parenchyma versus normal lumen (indicating the overall sacrifice in the black-blood contrast caused by DANTE), respectively.

For the patient group, 2 radiologists with 13 years (Q.Y.) and 8 years (G.W.) of experience in MR image interpretation, respectively, performed consensus reading. Combined conventional sequences from the 13 subjects were blindly reviewed for segment-level diagnosis of CVT. One month later, diagnostic review was also performed on the randomized 26 BTI image sets (2 sets per subject). Visible signals within the dark sinus lumens were deemed “apparent thrombi,” unless they were of granule shape (highly indicative of arachnoid granulations) or appeared as floating patches (highly indicative of residual blood rather than thrombi). Furthermore, the total volume of apparent thrombi was quantified in the superior sagittal sinus by manually contouring the thrombi section by section using commercial software (Vessel Analysis, Beijing Sirui Star Technology Co., Ltd.).

The 2 BTI image sets from each patient were then reviewed side by side to scrutinize the actual incidence of residual blood and true thrombus by the same 2 radiologists. The finalized true thrombi were individually categorized into hyperintense, isointense, or hybrid types on the basis of their appearance with respect to adjacent normal brain parenchyma. Signal intensity measurement was performed, respectively, for the regions of residual blood and residual blood nearest isointense and hyperintense thrombus, normal lumen, and brain parenchyma on both image sets. The CNRs of residual blood or thrombus to other tissues were calculated.

Statistical Analysis

Statistical analysis was performed using SPSS (Version 16.0; IBM, Armonk, New York). A 2-tailed Wilcoxon signed rank test was used to determine the difference in the thrombus volume, SNR, and CNR between the 2 BTI techniques. A Cohen κ test was used to determine the agreement in the diagnosis of CVT at the per-segment level between BTI and conventional imaging techniques. Conventional imaging techniques were used as the reference standard for assessing the sensitivity, specificity, and negative and positive predictive values of BTI techniques. Statistical significance was defined as P < .05.