Body composition assessment: comparison of quantitative values between magnetic resonance imaging and computed tomography

Chiara Zaffina; Rolf Wyttenbach; Alberto Pagnamenta; Rosario Francesco Grasso; Matteo Biroli; Filippo Del Grande; Stefania Rizzo

doi:10.21037/qims-21-619

Body composition assessment: comparison of quantitative values between magnetic resonance imaging and computed tomography

Quant Imaging Med Surg. 2022 Feb;12(2):1450-1466. doi: 10.21037/qims-21-619.

Authors

Chiara Zaffina¹, Rolf Wyttenbach^{2

3

4}, Alberto Pagnamenta^{5

6

7}, Rosario Francesco Grasso⁸, Matteo Biroli⁹, Filippo Del Grande^{2

3}, Stefania Rizzo^{2

3}

Affiliations

¹ Policlinico Universitario Campus Bio-Medico di Roma, Roma, Italy.
² Istituto di Imaging della Svizzera Italiana (IIMSI), Ente Ospedaliero Cantonale, Lugano, Switzerland.
³ Facoltà di Scienze Biomediche, Università della Svizzera Italiana, Lugano, Switzerland.
⁴ Department of Diagnostic, Interventional and Pediatric Radiology (DIPR), Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland.
⁵ Clinical Trial Unit, Ente Ospedaliero Cantonale, Lugano, Switzerland.
⁶ Intensive Care Unit, Ente Ospedaliero Cantonale, Mendrisio, Switzerland.
⁷ Division of Pneumology, University of Geneva, Geneva, Switzerland.
⁸ Departmental Faculty of Medicine and Surgery, Unit of Interventional Radiology, Università Campus Bio-Medico di Roma, Rome, Italy.
⁹ Humanitas University, Department of Biomedical Sciences, Pieve Emanuele, Milan, Italy.

Abstract

Background: The primary objective of this study was to compare measurements of skeletal muscle index (SMI), visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) at the level of L3, on subjects who underwent computed tomography (CT) and magnetic resonance imaging (MRI) examinations within a three-month period. The secondary objective was to compare the automatic and semi-automatic quantifications of the same values for CT images.

Methods: Among subjects who underwent CT and MRI at our Institution between 2011 and 2020, exclusion criteria were: presence of extensive artifacts; images not including the whole waist circumference; CT acquired with low-dose technique and lack of non-contrast images. A set of three axial images (CT, MRI T1-weighted and T2-weighted) were used to extract the following measurements with semi-automatic segmentations: SMI [calculated normalizing skeletal muscle area (SMA) by the square height], SAT, VAT. For the CT images only, the same values were also calculated by using automatic segmentation. Statistical analysis was performed comparing quantitative MRI and CT measurements by Pearson correlation analysis and by Bland-Altman agreement analysis.

Results: A total of 123 patients were included. By performing linear regression analysis, CT and MRI measurements of SMI showed a high correlation (r²=0.81 for T1, r²=0.89 for T2), with a mean logarithmic difference between CT and MRI quantitative values of 0.041 for T1-weighted and 0.072 for T2-weighted images. CT and MRI measurements of SAT showed high correlation (r²=0.81 for T1; r²=0.81 for T2), with a mean logarithmic difference between CT and MRI values of 0.0174 for T1-weighted and 0.201 for T2-weighted images. CT and MRI measurements of VAT showed high correlation (r²=0.94 for T1; r²=0.93 for T2), with a mean logarithmic difference of 0.040 for T1-weighted and -0.084 for T2-weighted images. The comparison of values extracted by semi-automatic and automatic segmentations were highly correlated.

Conclusions: Quantification of body composition values at MRI from T1-weighted and T2-weighted images was highly correlated to same values at CT, therefore quantitative values of body composition among patients who underwent either one of the examinations may be compared. CT body composition values extracted by semi-automatic and automatic segmentations showed high correlation.

Keywords: Body composition; computed tomography (CT); magnetic resonance imaging (MRI).