Standardized quantitative measurements of wrist cartilage in healthy humans using 3T magnetic resonance imaging

Abstract

Aim: To quantify the wrist cartilage cross-sectional area in humans from a 3D magnetic resonance imaging (MRI) dataset and to assess the corresponding reproducibility.

Methods: The study was conducted in 14 healthy volunteers (6 females and 8 males) between 30 and 58 years old and devoid of articular pain. Subjects were asked to lie down in the supine position with the right hand positioned above the pelvic region on top of a home-built rigid platform attached to the scanner bed. The wrist was wrapped with a flexible surface coil. MRI investigations were performed at 3T (Verio-Siemens) using volume interpolated breath hold examination (VIBE) and dual echo steady state (DESS) MRI sequences. Cartilage cross sectional area (CSA) was measured on a slice of interest selected from a 3D dataset of the entire carpus and metacarpal-phalangeal areas on the basis of anatomical criteria using conventional image processing radiology software. Cartilage cross-sectional areas between opposite bones in the carpal region were manually selected and quantified using a thresholding method.

Results: Cartilage CSA measurements performed on a selected predefined slice were 292.4 ± 39 mm(2) using the VIBE sequence and slightly lower, 270.4 ± 50.6 mm(2), with the DESS sequence. The inter (14.1%) and intra (2.4%) subject variability was similar for both MRI methods. The coefficients of variation computed for the repeated measurements were also comparable for the VIBE (2.4%) and the DESS (4.8%) sequences. The carpus length averaged over the group was 37.5 ± 2.8 mm with a 7.45% between-subjects coefficient of variation. Of note, wrist cartilage CSA measured with either the VIBE or the DESS sequences was linearly related to the carpal bone length. The variability between subjects was significantly reduced to 8.4% when the CSA was normalized with respect to the carpal bone length.

Conclusion: The ratio between wrist cartilage CSA and carpal bone length is a highly reproducible standardized measurement which normalizes the natural diversity between individuals.

Keywords: Cartilage; Magnetic resonance imaging; Quantification; Wrist.

Figure 1

Standardized selection of a slice of interest. A: Axial slice of the wrist in which we initially determined the axis of the anterior part of the Radius (1) (green line) and then the perpendicular sagittal axis (orange line); B: Corresponding sagittal slice in which, we chose the slice going through the proximal part of the capitate (3) and the metacarpal basis (6); C: Corresponding coronal slice showing the radius (1), ulna (5), navicular (4), semi-lunar (2) and capitate bones (3). Arrow indicates the carpal bone length measurement from the lowest point of the capitate to the highest point of the semi-lunar bone.

Figure 2

Manual segmentation of the cartilage cross-sectional area. A: Coronal slice (as in Figure 1C) showing different joints between radius and scaphoid (1), radius and lunate (2), lunate and ulna (3), scaphoid and capitate (4), lunate and ulna (5), lunate and triquetrum (6), triquetrum and hamate (7) hamate and capitate (8), carpals bones and metacarpal bones (9) scaphoid and lunate (10); B: Coronal slice illustrating the manual segmentation of the cartilage area of interest and the corresponding result (C).

Figure 3

Relationships between wrist cartilage cross-sectional area measurements. Relationship between wrist cartilage CSA (mm²) and carpal bone length (mm) measured. The cartilage CSA was measured using the VIBE (A) and the DESS (B) MRI sequence; C: Relationship between cartilage CSA measured using the VIBE (x-axis) and the DESS (y-axis) MRI sequences. The continuous line crossing the both axes represents the identity line. VIBE: Volume interpolated breath hold examination; DESS: Dual echo steady state; CSA: Cartilage cross-sectional area; MRI: Magnetic resonance imaging.