Figure 1.
Pulse sequence for 2D IR-dUTE imaging of the ultrashort T2* components in white matter of the brain.
A minimal nominal TE of 8 µs is achieved by using half pulse excitation and radial ramp sampling. An adiabatic inversion recovery preparation pulse is used together with dual echo acquisition to suppress the signal from the long T2 components in white matter and create high contrast for the ultrashort T2* components.
Figure 2.
Non-selective 2D UTE measurement of T2* (A) and T1 (B, C) of a rubber phantom.
A short T2* of 322±4 µs was measured by exponential decay curve fitting of UTE images at a series of different TEs (A). A short T1 of 182±6 ms was measured by exponential recovery curve fitting of saturation recovery UTE images acquired at a series of saturation recovery times (B). The novel IR-dUTE approach gave a short T1 of 198±8 ms (C), which is only a 3% overestimate relative to (B), suggesting that the new IR based technique is capable of accurate T1 measurement of short T2* species.
Figure 3.
Non-selective 2D UTE measurement of T1 of a piece of bovine cortical bone: a short T1 of 253±18 ms was measured by exponential recovery curve fitting of saturation recovery UTE images acquired at a series of saturation recovery times (A).
The novel IR-dUTE approach gave a short T1 of 266±9 ms (B), which is only a 5% overestimate relative to (A), further suggesting that the new IR based technique is capable of accurate T1 measurement of short T2* species such as cortical bone.
Figure 4.
Conventional clinical imaging of the brain of a healthy volunteer using T1-FSE (A), T2-FSE (B), PD-FDE (C), GRE (D), T2-FLAIR (E), MP-RAGE (F), and IR-dUTE imaging with a TE of 8 µs (G) and 2.2 ms (H) as well as the corresponding echo subtraction image (I).
The ultrashort T2* components in the white matter of the centrum semiovale appear hypointense in the 1st echo image (G) but near zero signal in the 2nd echo image (H), and are highlighted in the subtraction image (I). The rubber phantom is only seen with UTE based techniques.
Figure 5.
Selected 2D IR-dUTE imaging of a 24 year old healthy volunteer with four sets of dual-echo acquisitions, including TEs of 8 µs (A) and 2.2 ms (E), 0.1 ms (B) and 2.2 ms (F), 0.4 ms (C) and 2.2 ms (G), 2.2 ms (D) and 4.4 ms (H) as well as the corresponding echo subtraction images (I–L).
Single component fitting shows a T2* of 284±46 µs for the ultrashort T2* components in white matter of this volunteer. The rubber phantom showed a short T2* of 341±25 µs, which is only a 5.9% overestimate of the true T2* value of 322±9 µs derived from the non-selective UTE acquisitions.
Figure 6.
Selected 2D IR-dUTE imaging of a 27 year old healthy volunteer with four different TR/TI combinations: 750/260 ms (A, E, I), 1000/320 ms (B, F, J), 1500/420 ms (C, G, K) and 2000/480 ms (D, H, L).
The ultrashort T2* components were depicted with low signal in the first echo images (A–D), and near zero signal in the 2nd echo images (E–H), but highlighted in the echo subtraction images (I–L). Exponential recovery curve fitting shows a short T1 of 195±14 ms for the rubber, which is about a 7% overestimate relative to the reference value of 182 ms derived from the non-selective saturation recovery UTE acquisitions. A short T1 of 234±80 was demonstrated for the ultrashort T2* components in white matter. A significant fitting standard error was observed.