Spatial patterns of brain lesions assessed through covariance estimations of lesional voxels in multiple Sclerosis: The SPACE-MS technique

Abstract

Predicting disability in progressive multiple sclerosis (MS) is extremely challenging. Although there is some evidence that the spatial distribution of white matter (WM) lesions may play a role in disability accumulation, the lack of well-established quantitative metrics that characterise these aspects of MS pathology makes it difficult to assess their relevance for clinical progression. This study introduces a novel approach, called SPACE-MS, to quantitatively characterise spatial distributional features of brain MS lesions, so that these can be assessed as predictors of disability accumulation. In SPACE-MS, the covariance matrix of the spatial positions of each patient's lesional voxels is computed and its eigenvalues extracted. These are combined to derive rotationally-invariant metrics known to be common and robust descriptors of ellipsoid shape such as anisotropy, planarity and sphericity. Additionally, SPACE-MS metrics include a neuraxis caudality index, which we defined for the whole-brain lesion mask as well as for the most caudal brain lesion. These indicate how distant from the supplementary motor cortex (along the neuraxis) the whole-brain mask or the most caudal brain lesions are. We applied SPACE-MS to data from 515 patients involved in three studies: the MS-SMART (NCT01910259) and MS-STAT1 (NCT00647348) secondary progressive MS trials, and an observational study of primary and secondary progressive MS. Patients were assessed on motor and cognitive disability scales and underwent structural brain MRI (1.5/3.0 T), at baseline and after 2 years. The MRI protocol included 3DT1-weighted (1x1x1mm³) and 2DT2-weighted (1x1x3mm³) anatomical imaging. WM lesions were semiautomatically segmented on the T2-weighted scans, deriving whole-brain lesion masks. After co-registering the masks to the T1 images, SPACE-MS metrics were calculated and analysed through a series of multiple linear regression models, which were built to assess the ability of spatial distributional metrics to explain concurrent and future disability after adjusting for confounders. Patients whose WM lesions laid more caudally along the neuraxis or were more isotropically distributed in the brain (i.e. with whole-brain lesion masks displaying a high sphericity index) at baseline had greater motor and/or cognitive disability at baseline and over time, independently of brain lesion load and atrophy measures. In conclusion, here we introduced the SPACE-MS approach, which we showed is able to capture clinically relevant spatial distributional features of MS lesions independently of the sheer amount of lesions and brain tissue loss. Location of lesions in lower parts of the brain, where neurite density is particularly high, such as in the cerebellum and brainstem, and greater spatial spreading of lesions (i.e. more isotropic whole-brain lesion masks), possibly reflecting a higher number of WM tracts involved, are associated with clinical deterioration in progressive MS. The usefulness of the SPACE-MS approach, here demonstrated in MS, may be explored in other conditions also characterised by the presence of brain WM lesions.

Keywords: Anisotropy; Caudality; Lesion spatial distribution; Magnetic resonance imaging; Multiple sclerosis; SPACE-MS.

Graphical abstract

Fig. 1

Analysis pipeline A. Once all the individual lesions were defined, the 3D positions of all lesional voxels within the whole-brain mask were extracted. These 3D positions were used to evaluate their spatial covariance matrix, from which the eigenvalues were computed after performing a principal component analysis. The eigenvalues were then used to obtain some of the spatial distribution metrics (see Methods section). B. This figure shows the definition of neuraxis and of neuraxis caudality index (NCI) for a given lesion (lesion NCI). The whole-brain NCI was computed in the same way, substituting the CoM of each individual lesion for the CoM of the whole-brain lesion mask. The maximum lesion NCI was computed calculating the CoM of the lowermost brain lesion. Abbreviations (in alphabetical order): CAI: Covariance Anisotropy Index; CoM: centre of mass; CPI: Covariance Planarity Index; CSI: Covariance Sphericity Index; MCI: Mean Covariance Index; NCI: neuraxis caudality index; SMA: supplementary motor area.

Fig. 2

Examples of spatial distribution metrics. This figure shows the significance of our spatial distribution metrics. In A, there is an example of a lesion mask with low NCI (whose CoM lies very close to the SMA) and another example of a lesion mask with very high NCI (whose CoM lies very close to the brainstem). In B, there are schematic descriptions of different types of lesion masks, with some real examples: on the left, a whole-brain lesion mask with a very high CAI, where lesions would spread anisotropically, mainly across one spatial direction; in the middle, a lesion mask with a very high CPI, where lesions would spread mainly across two directions; on the right, a lesion mask with a very high CSI, where lesions would spread isotropically, i.e. across all three spatial directions almost equally. Abbreviations: CAI: covariance anisotropy index; CPI: covariance planarity index; CSI: covariance sphericity index; NCI: neuraxis caudality index.

Fig. 3

Lesion detection across time points. This figure shows how individual lesions were defined on the whole-brain lesion mask. Independent connected components were detected in a lesion mask obtained by merging the lesions masks corresponding from different time points (referred to as all-time-points merged image).

Fig. 4

Adjusted values of SPACE-MS metrics for all study centres. This figure shows the age-, sex-, disease-duration-, and lesion-load-adjusted values of SPACE-MS metrics for all 16 study centres. Overall, such adjusted values of SPACE-MS metrics were very similar across centres. Importantly, though, the values shown in this figure have not been adjusted for disability measures, so the variability across centres due to differences in disability scores between centres has not been removed. Abbreviations: CAI: covariance anisotropy index; CPI: covariance planarity index; CSI: covariance sphericity index; NCI: neuraxis caudality index.

Fig. 5

Baseline associations between spatial distribution metrics and clinical variables. This figure shows the main associations between spatial distribution metrics and clinical variables at baseline, after adjusting for all relevant confounders. For graphical purposes only, the y-axis in this figure indicates the values of the clinical variable at baseline (dependent variable) after having been adjusted for all the covariates in the corresponding ‘best a priori model’ (as indicated in the methods and Table 3). Abbreviations in alphabetical order: CSI: covariance sphericity index; EDSS: expanded disability status scale; NCI: neuraxis caudality index; NHPT: nine-hole peg test; PASAT: paced auditory serial addition test; SDMT: symbol digit modalities test; TWT: 25-foot timed walk test.