EVALUATING COMPRESSED SENSE ACCELERATION FOR QUANTITATIVE MAPPING OF LONGITUDINAL RELAXATION RATE R1

Hintergrund: Absolute measurements of the longitudinal relaxation rate R1 (= 1/T1) have a high potential for diagnostic applications because they are more sensitive to pathological micro-structural changes from tumors or diseases like multiple sclerosis (1) compared to conventional R1w or T1w imaging. Moreover, measured values are assumed to be sequence and hardware independent (2-3) and function as more specific markers for disease progression and contrast agent uptake (1). We investigated the effect of Compressed SENSE (4) (CS) acceleration on the fidelity of R1 mapping to foster clinical applicability by decreasing scan time. Methoden: Five healthy subjects were scanned on a Philips 3T Elition. The MRI protocol consisted of B1 mapping and two T1w sequences with different flip angles. Measurements were performed with standard SENSE (S = 2.5, scan time: 11 min) and 2 different CS acceleration factors CS = 4 (9.5 min) and CS = 6 (6 min). R1 maps were computed using the hMRI toolbox (5-6). Gray matter (GM) and white matter (WM) segments were obtained using SPM12 (7). Ergebnisse: Visually, R1 maps obtained with different accelerations are very similar (Fig1). Slight artifacts from standard SENSE reconstruction (arrow Fig1a) were eliminated by CS (Fig1b-c). Subject average R1 values in GM, WM, and manually defined VOIs (Fig1d) agree well between the 3 protocols (Fig2). The largest variance of R1 values across subjects is found with SENSE, while CS reduces R1 value variations across subjects (Fig2). Both CS protocols show comparable results. Diskussion: Our comparison of R1 mapping protocols with different accelerations reveals R1 maps with comparable quality. Quantitative R1 values depend neither on acceleration technique (SENSE vs. CS) nor factor (CS = 4 vs. CS = 6). Our results suggest that compressed SENSE, with acceleration factors up to at least 6, can be used for quantitative R1 mapping without loss of fidelity but with the advantage of clearly reduced scan times. Fazit: CS is highly promising to establish quantitative R1 mapping within clinically feasible scan times.