Data‐Driven Quantitative Susceptibility Mapping Using Loss Adaptive Dipole Inversion (LADI)

• Published in: Journal of magnetic resonance imaging 2020-09, Vol.52 (3), p.823-835
• Main Authors: Kamesh Iyer, Srikant, Moon, Brianna F., Josselyn, Nicholas, Ruparel, Kosha, Roalf, David, Song, Jae W., Guiry, Samantha, Ware, Jeffrey B., Kurtz, Robert M., Chawla, Sanjeev, Nabavizadeh, S. Ali, Witschey, Walter R.
• Format: Article
• Language: English
• Subjects: Algorithms; Animal models; Animals; Bias; Brain - diagnostic imaging; Brain Mapping; compressed sensing; Cosmos; dipole inversion; Dipoles; Error analysis; Field strength; Glioblastoma; Humans; Image degradation; Image processing; Image Processing, Computer-Assisted; Image quality; Image reconstruction; Inversion; iron; Magnetic Resonance Imaging; Mapping; Medical imaging; Morphology; Neuroimaging; Quality assessment; Quantitative susceptibility mapping; Retrospective Studies; Root-mean-square errors; Scanners; Sharpness; Statistical analysis; Statistical tests; Susceptibility; Swine
• ISSN: 1053-1807; 1522-2586
• DOI: 10.1002/jmri.27103

Background Quantitative susceptibility mapping (QSM) uses prior information to reconstruct maps, but prior information may not show pathology and introduce inconsistencies with susceptibility maps, degrade image quality and inadvertently smoothing image features. Purpose To develop a local field data‐driven QSM reconstruction that does not depend on spatial edge prior information. Study Type Retrospective. Subjects, animal models A dataset from 2016 ISMRM QSM Challenge, 11 patients with glioblastoma, a patient with microbleeds and porcine heart. Sequence/Field Strength 3D gradient echo sequence on 3T and 7T scanners. Assessment Accuracy was compared to Calculation of Susceptibility through Multiple Orientation Sampling (COSMOS), and several published techniques using region of interest (ROI) measurements, root‐mean‐squared error (RMSE), structural similarity index metric (SSIM), and high‐frequency error norm (HFEN). Numerical ranking and semiquantitative image grading was performed by three expert observers to assess overall image quality (IQ) and image sharpness (IS). Statistical Tests Bland–Altman, Friedman test, and Conover multiple comparisons. Results Loss adaptive dipole inversion (LADI) (β = 0.82, R2 = 0.96), morphology‐enabled dipole inversion (MEDI) (β = 0.91, R2 = 0.97), and fast nonlinear susceptibility inversion (FANSI) (β = 0.81, R2 = 0.98) had excellent correlation with COSMOS and no bias was detected (bias = 0.006 ± 0.014, P < 0.05). In glioblastoma patients, LADI showed consistently better performance (IQGrade = 2.6 ± 0.4, ISGrade = 2.6 ± 0.3, IQRank = 3.5 ± 0.4, ISRank = 3.9 ± 0.2) compared with MEDI (IQGrade = 2.1 ± 0.3, ISGrade = 2 ± 0.5, IQRank = 2.4 ± 0.5, ISRank = 2.8 ± 0.2) and FANSI (IQGrade = 2.2 ± 0.5, ISGrade = 2 ± 0.4, IQRank = 2.8 ± 0.3, ISRank = 2.1 ± 0.2). Dark artifact visible near the infarcted region in MEDI (InfMEDI = −0.27 ± 0.06 ppm) was better mitigated by FANSI (InfFANSI‐TGV = −0.17 ± 0.05 ppm) and LADI (InfLADI = −0.18 ± 0.05 ppm). Conclusion For neuroimaging applications, LADI preserved image sharpness and fine features in glioblastoma and microbleed patients. LADI performed better at mitigating artifacts in cardiac QSM. Evidence Level 4 Technical Efficacy Stage 1 J. Magn. Reson. Imaging 2020;52:823–835.