Double spectral attenuated inversion recovery (DSPAIR)—an efficient fat suppression technique for late gadolinium enhancement at 3 tesla

• Published in: NMR in biomedicine 2021-10, Vol.34 (10), p.e4580-n/a
• Main Authors: Jenista, Elizabeth R., Jensen, Christoph J., Wendell, David, Spatz, Deneen, Darty, Stephen, Kim, Han W., Parker, Michele, Klem, Igor, Chen, Enn‐Ling, Kim, Raymond J., Rehwald, Wolfgang G.
• Format: Article
• Language: English
• Subjects: Arrhythmogenic Right Ventricular Dysplasia - diagnostic imaging; Computer Simulation; fat suppression (FS); Gadolinium - chemistry; Humans; inversion recovery (IR); late gadolinium enhancement (LGE); Lipids - chemistry; Magnetic Resonance Imaging; Middle Aged; myocardial infarction (MI); Phantoms, Imaging; phase sensitive inversion recovery (PSIR); Signal-To-Noise Ratio; spectral attenuated inversion recovery (SPAIR); spectral presaturation with inversion recovery (SPIR)
• ISSN: 0952-3480; 1099-1492
• DOI: 10.1002/nbm.4580

Despite clinical use of late gadolinium enhancement (LGE) for two decades, an efficient, robust fat suppression (FS) technique still does not exist for this CMR mainstay. In ischemic and non‐ischemic heart disease, differentiating fibrotic tissue from infiltrating and adjacent fat is crucial. Multiple groups have independently developed an FS technique for LGE, double spectral attenuated inversion recovery (DSPAIR), but no comprehensive evaluation was performed. This study aims to fill this gap. DSPAIR uses two SPAIR pulses and one non‐selective IR pulse to enable FS LGE, including compatibility with phase sensitive inversion recovery (PSIR). We implemented a magnitude (MAGN) and a PSIR variant and compared them with LGE without FS (CONTROL) and with spectral presaturation with inversion recovery (SPIR) in simulations, phantoms, and patients. Fat magnetization by SPIR, MAGN DSPAIR, and PSIR DSPAIR was simulated as a function of pulse B1, readout (RO) pulse number, and fat TI. A phantom with fat, fibrosis, and myocardium compartments was imaged using all FS methods and modifying pulse B1, RO pulse number, and heart rate. Signal was measured in SNR units. Fat, myocardium, and fibrosis SNR and fibrosis‐to‐fat CNR were obtained. Patient images were acquired with all FS techniques. Fat, myocardium, and fibrosis SNR, fibrosis‐to‐fat CNR, and image and FS quality were assessed. In the phantom, both DSPAIR variants provided superior FS compared with SPIR, independent of heart rate and RO pulse number. MAGN DSPAIR reduced fat signal by 99% compared with CONTROL, PSIR DSPAIR by 116%, and SPIR by 67% (25 RO pulses). In patients, both DSPAIR variants substantially reduced fat signal (MAGN DSPAIR by 87.1% ± 10.0%, PSIR DSPAIR by 130.5% ± 36.3%), but SPIR did not (35.8% ± 25.5%). FS quality was good to excellent for MAGN and PSIR DSPAIR, and moderate to poor for SPIR. DSPAIR provided highly effective FS across a wide range of parameters. PSIR DSPAIR performed best. Double spectral attenuated inversion recovery (DSPAIR) provided highly effective fat suppression for late gadolinium enhancement across a wide range of parameters while offering compatibility with phase sensitive inversion recovery (PSIR). The DSPAIR magnitude and PSIR variants outperformed the common spectral presaturation with inversion recovery (SPIR) technique regarding fat signal reduction and fibrosis‐fat CNR, in simulations, phantoms, and patients. PSIR DSPAIR scored best, with a 131% fat signal reduct