Prospective comparison of brain tumor detection and characterization using post-contrast Dixon uT1RESS and MPRAGE at 3 Tesla

Abstract

Purpose

This study aimed to evaluate the image quality and lesion characterization with Dixon unbalanced T1 relaxation-enahnced steady-state (uT1RESS) in comparison to magnetization-prepared rapid acquisition gradient echo (MPRAGE) pulse sequence for detecting brain tumors, focusing on its potential to improve diagnostic accuracy and enhance brain tumor assessment.

Methods

This single-center prospective study enrolled 20 patients (12 males, aged 35–83), with primary brain tumors and brain metastases. Both MPRAGE and Dixon uT1RESS were acquired at 3 T. Objective image quality was assessed by contrast-to-noise ratio (CNR), tumor-to-brain contrast, and tumor volume. Subjective image quality was assessed by two independent readers focusing on lesion visibility, lesion margins, motion, and static artifacts. A side-by-side comparison assessed diagnostic performance regarding lesion detection, evaluation of internal structure and vascular or dural invasion.

Results

Dixon uT1RESS had a reduced acquisition time (2 min 51 s vs. 4 min 52 s for MPRAGE) and showed significantly higher CNR and tumor-to-brain contrast compared to MPRAGE (p < 0.001). Subjectively, both sequences showed similar overall image quality. Dixon uT1RESS achieved more conspicuous lesions with better-defined lesion margins (p < 0.001), while MPRAGE performed better in evaluating internal lesion structure (p < 0.05). Dixon uT1RESS was rated better for lesion detection, with three lesions additionally identified on this sequence. Wilcoxon signed-rank test was used to assess differences.

Conclusions

Dixon uT1RESS significantly improved tumor conspicuity and detection, particularly for small metastatic lesions. This new technique offers promising potential for enhancing clinical brain tumor imaging, though additional sequences may be necessary for comprehensive lesion characterization.