Comparison of quantitative measurement of signal intensity of nucleus pulposus with visual interpretation as method for classification of disc degeneration     — The International Society for the Study of the Lumbar Spine
  1. Schedule
  2. GENERAL POSTER SESSION
  3. Comparison of quantitative measurement of signal intensity of nucleus pulposus with visual interpretation as method for classification of disc degeneration    

Comparison of quantitative measurement of signal intensity of nucleus pulposus with visual interpretation as method for classification of disc degeneration     (#1047)

Niko Murto 1 , Liisa Kerttula 1 , Teija Lund 2 , Katariina Luoma 1 3
  1. Department of Radiology, Helsinki University Central Hospital, Helsinki, Finland
  2. Töölö Hospital, Helsinki University Central Hospital, Helsinki, Finland
  3. University of Helsinki, Helsinki, Finland

 

Introduction:

Automated quantitative methods for evaluation and grading of disc degeneration (DD) on MRI are needed for AI-based diagnosis of low back pain1. Decrease of signal intensity (SI) of the intervertebral disc (IVD) on T2-weighted MRI (T2WI) is considered as sign of DD as it correlates with loss of proteoglycan and water content in nucleus pulposus (NP)2. NP SI is commonly evaluated with visual grading3, but subjectivity and categorization reduce reliability of these grading systems. Quantitative measurement methods of NP SI have been published but variable regions of the IVD such as collagenous annulus fibrosus and fibrous horizontal band, intranuclear cleft (INC) in the middle of the NP are included in these measurements. These quantitative methods have excellent reliability4–7. However, far too little attention has been paid to comparison between different measurement methods8. Our aim is to find a reliable method for DD evaluation by comparing different quantitative measurement methods of NP SI with visual interpretation.

Methods:

On lumbar MRI of 19 men (mean age 51 years, SD 0.7), mean SI of 95 IVDs was measured from mid-sagittal T2WI by two observers using three regions of interest (ROI) with different sizes: including the whole IVD (WD-ROI), ellipsoid on the NP including INC (ELLIPS-ROI) and targeted, freehand ROI on the most homogenous, bright area of the NP excluding INC (TARGET-ROI). Mean SI values were adjusted using SI of adjacent cerebrospinal fluid as reference to have comparable values between subjects. Observers graded NP SI also visually with 5-point grading system (Table 1) and assessed DD with Pfirrmann grading. Inter- and intra-observer reliability were assessed with intra-class correlation (ICC) for SI measurements and weighted kappa (κ) for visual grading. Correlation of SI measurements with visual grading was assessed with Spearman's correlation (ρ).

Results:

Reliability of SI measurements was excellent (ICC > 0.94). Reliability of visually graded NP SI varied from moderate to substantial (κ = 0.56-0.72). TARGET-ROI illustrated most clearly the difference of measured NP SI values between visual grades of NP SI (Figure 1). A small group of obviously degenerated discs with widely decreased SI but a small region with high SI in NP (“zebra”) was distinct with TARGET-ROI but not with WD-ROI. SI measurements, excluding zebra discs, had a strong negative correlation with visually graded NP SI and Pfirrmann grading. Visually graded NP SI correlated strongest with TARGET-ROI (ρ = -0.84) while Pfirrmann grading strongest with ELLIPS-ROI (ρ = -0.83) and WD-ROI (p=-0.82).

Discussion:

Targeted SI measurement of the most homogeneous NP region gives the best tissue characterization of the NP. INC has a different SI than NP tissue above and below it because of fibrous tissue composition. TARGET-ROI showed most clearly the differences between visually graded NP SI categories. It also showed that some degenerated discs had an exceptionally high SI in NP. For reliable DD grading, we suggest excluding INC from SI measurements of NP as INC has a different tissue composition than the rest of NP region.

 

618f84eff1268-Table1.png

 

618f84eff1268-Figure1.png

 
  1. D’Antoni F, Russo F, Ambrosio L, et al. Artificial Intelligence and Computer Vision in Low Back Pain: A Systematic Review. International Journal of Environmental Research and Public Health 2021;18:10909.
  2. Tertti M, Paajanen H, Laato M, et al. Disc degeneration in magnetic resonance imaging. A comparative biochemical, histologic, and radiologic study in cadaver spines. Spine 1991;16:629–34.
  3. Pfirrmann CWA, Metzdorf A, Zanetti M, et al. Magnetic Resonance Classification of Lumbar Intervertebral Disc Degeneration. Spine 2001;26:1873–8.
  4. Videman T, Battié MC, Gibbons LE, et al. A new quantitative measure of disc degeneration. Spine Journal 2017;17:746–53.
  5. Salamat S, Hutchings J, Kwong C, et al. The relationship between quantitative measures of disc height and disc signal intensity with Pfirrmann score of disc degeneration. SpringerPlus;5. Epub ahead of print December 1, 2016. DOI: 10.1186/s40064-016-2542-5.
  6. Abdollah V, Parent EC, Battié MC. Is the location of the signal intensity weighted centroid a reliable measurement of fluid displacement within the disc? Biomedizinische Technik Biomedical engineering 2018;63:453–60.
  7. Stelzeneder D, Welsch GH, Kovács BK, et al. Quantitative T2 evaluation at 3.0 T compared to morphological grading of the lumbar intervertebral disc: A standardized evaluation approach in patients with low back pain. European Journal of Radiology 2012;81:324–30.
  8. Nagashima M, Abe H, Amaya K, et al. A method for quantifying intervertebral disc signal intensity on T2-weighted imaging. Acta Radiologica 2012;53:1059–65.
#ISSLS2022