High-contrast osteochondral junction imaging using a 3D dual adiabatic inversion recovery-prepared ultrashort echo time cones sequence.

2021 
While conventional MRI sequences cannot visualize tissues from the osteochondral junction (OCJ) due to these tissues' short transverse T2 /T2 * relaxations, ultrashort echo time (UTE) sequences can overcome this limitation. A 2D UTE sequence with a dual adiabatic inversion recovery preparation (DIR-UTE) for selective imaging of short T2 tissues with high contrast has previously been developed, but high sensitivity to eddy currents and aliased out-of-slice excitation make it difficult to image the thin layer of the OCJ in vivo. Here, we combine the DIR scheme with a 3D UTE cones sequence for volumetric imaging of OCJ tissues in vivo, aiming to generate higher OCJ contrast compared with a recently developed single IR-prepared UTE sequence with a fat saturation module (IR-FS-UTE). All sequences were implemented on a 3-T clinical scanner. The DIR-UTE cones sequence combined a 3D UTE cones sequence with two narrow-band adiabatic IR preparation pulses centered on water and fat spectrum frequencies, respectively. The 3D DIR-UTE cones sequence was first applied to a phantom, then to the knees of four healthy volunteers and four patients diagnosed with osteoarthritis and compared with the IR-FS-UTE sequence. In both phantom and volunteer studies, the proposed DIR-UTE cones sequence showed much higher contrast for OCJ imaging than the IR-FS-UTE sequence did. The 3D DIR-UTE cones sequence showed a significantly higher contrast-to-noise ratio between the OCJ and subchondral bone fat (mean, standard deviation [SD]: 25.7 ± 2.3) and between the OCJ and superficial layers of cartilage (mean, SD: 22.2 ± 3.5) compared with the IR-FS-UTE sequence (mean, SD: 10.8 ± 2.5 and 16.3 ± 2.6, respectively). The 3D DIR-UTE cones sequence is feasible for imaging of the OCJ region of the knee in vivo and produces both high resolution and high contrast.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    33
    References
    0
    Citations
    NaN
    KQI
    []