ADC Quantification of the Vertebral Bone Marrow Water Component: Removing the Confounding Effect of Residual Fat

• Published in: Magnetic resonance in medicine Vol. 78; no. 4; pp. 1432 - 1441
• Main Authors: Dieckmeyer, Michael, Ruschke, Stefan, Eggers, Holger, Kooijman, Hendrik, Rummeny, Ernst J., Kirschke, Jan S., Baum, Thomas, Karampinos, Dimitrios C.
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.10.2017
• Subjects: ADC quantification; Adipose Tissue - diagnostic imaging; Adult; Bias; Bone imaging; Bone marrow; Bone Marrow - diagnostic imaging; Diffusion coefficient; Diffusion Magnetic Resonance Imaging - methods; diffusion‐weighted imaging (DWI); Female; Humans; Image Processing, Computer-Assisted - methods; Imaging; Magnetic resonance spectroscopy; magnetic resonance spectroscopy (MRS); Magnetic Resonance Spectroscopy - methods; Male; Proton density (concentration); Spectroscopy; spine; Spine - diagnostic imaging; Vertebrae; Young Adult; diffusion-weighted imaging (DWI); spine; magnetic resonance spectroscopy (MRS); bone marrow; ADC quantification
• ISSN: 0740-3194; 1522-2594; 1522-2594
• DOI: 10.1002/mrm.26550

Purpose To remove the confounding effect of unsuppressed fat on the imaging‐based apparent diffusion coefficient (ADC) of the vertebral bone marrow water component when using spectrally selective fat suppression and to compare and validate the proposed quantification strategy against diffusion‐weighted magnetic resonance spectroscopy (DW‐MRS). Methods Twelve subjects underwent diffusion‐weighted imaging (DWI) and DW‐MRS of the vertebral bone marrow. A theoretical model was developed to take into account and correct the effects of residual fat on ADC, incorporating additional measurements for proton density fat fraction (PDFF) and water T2 (T2w). Uncorrected and corrected DWI‐based ADC was compared with DW‐MRS–based ADC using the Bland–Altman method. Results There was a systematic bias equal to 0.118 ± 0.116 × 10−3 mm2/s between DWI and DW‐MRS when no correction was performed. Taking into account measured PDFF and constant T2w reduced the bias to 0.006 ± 0.128 × 10−3 mm2/s. Using the proposed approach with both individually measured PDFF and T2w reduced both the bias and the limits of agreement between DWI and DW‐MRS (0.018 ± 0.065 × 10−3 mm2/s). Conclusion By taking into account the presence of residual fat in a modified signal model that incorporates additional individual measurements of PDFF and T2w, good agreement of imaging‐based ADC with MRS‐based ADC can be achieved in vertebral bone marrow. Magn Reson Med 78:1432–1441, 2017. © 2016 International Society for Magnetic Resonance in Medicine.