Background

A deep-learning prototype method, called CVSNet, was recently introduced for the automated detection of the central vein sign (CVS) in brain lesions and demonstrated effective and accurate discrimination of multiple sclerosis (MS) from its mimics. However, this method solely considered focal lesions displaying the central vein sign (CVS+) or not (CVS−), therefore requiring a manual pre-selection of the lesions to be evaluated by eliminating the so-called excluded lesions (CVSe) as defined by the NAIMS criteria. CVSe lesions may however play an important role in differential diagnosis. Moreover, extending the automated CVS classification to these lesions would facilitate the integration of CVSNet with existing MS lesion segmentation algorithms in a fully automated pipeline.

Objectives

To develop an improved version of the CVSNet prototype method able to classify all types of lesions (CVS+, CVS− and CVSe).

Methods

Patients with an established MS or CIS diagnosis (RRMS 29; SPMS 10; PPMS 10; CIS 1; mean ± SD age: 50 ± 11 years; male/female: 23/27), and healthy controls (n=8; mean ± SD age: 41 ± 9 years; male/female: 5/3), underwent 3T brain MRI (MAGNETOM Skyra and MAGNETOM Prisma, Siemens Healthcare, Erlangen, Germany, or Achieva, Philips Healthcare, Best, Netherlands). Brain lesions were automatically segmented and manually corrected by a single rater. CVS assessment was conducted on FLAIR* images by two raters, according to the NAIMS guidelines, yielding 1542 CVS+, 1004 CVS−, and 1131 CVSe lesions. A convolutional neural network (CNN) based on the CVSnet architecture was trained with different configurations using 3021 samples (1261 CVS+, 847 CVS−, and 913 CVSe) and evaluated in 656 unseen samples (281 CVS+, 157 CVS−, and 218 CVSe, from 13 patients) for final testing. The configurations relied on different combinations of the following channels as input: (i) FLAIR*, (ii) T2*, (iii) lesion mask, and (iv) CSF and brain tissue concentration maps obtained from a partial-volume estimation algorithm. Lesion-wise classification performance was evaluated for the different configurations by estimating the sensitivity, specificity, and accuracy for each lesion class.

Results

The results were similar across the different configurations. The best performance in the unseen testing set was obtained when all channels were used as input (sensitivity: 0.71, 0.73; specificity: 0.71, 0.81; and accuracy: 0.71, 0.79 for CVS+, CVS−, respectively). For CVSe, this approach achieved 0.52 sensitivity, 0.94 specificity, and 0.80 accuracy.

Conclusions

We introduced a modified CVSNet prototype method that can analyze the presence of the central vein for all types of brain lesions, enabling its integration with current MS lesion segmentation algorithms. This new feature will allow a fully automated assessment of the CVS in patients’ brains, speeding up the evaluation of CVS as a diagnostic biomarker for differentiating MS from mimicking diseases.

Background

Recent studies have described Multiple Sclerosis (MS) as a disconnection syndrome (Rocca et al. 2015). Modelling disconnectomes using brain networks enables to quantify connectivity loss using graph analysis. To build structural connectomes, high-quality diffusion Magnetic Resonance Imaging (dMRI) and robust tractography algorithms are typically required. However, high-quality dMRI is rarely acquired in clinical workups due to time constraints.

Objectives

We propose to use a tractography atlas to extract brain connectivity loss in response to lesions without requiring dMRI, and to model structural disconnectomes with brain graphs. Topological graph features are proposed as new radiological biomarkers and their relation with Total Lesion Volume (TLV) and Expanded Disability Status Scale (EDSS) are studied.

Methods

589 MS patients (159 males, age 28±8yo, EDSS 2.40±1.22, TLV 13.0±14.6mL) underwent MRI at 3T (MAGNETOM Skyra, Siemens Healthcare, Erlangen, Germany). Acquisition protocols included T1-weighted magnetization-prepared rapid gradient echo (MPRAGE) and fluid-attenuated inversion recovery (FLAIR).

Lesions were segmented using LeMan-PV, a prototype lesion segmentation algorithm (Fartaria et al. 2016). The lesion masks were registered to standard MNI space and overlapped with the HCP842 tractography atlas (Yeh et al. 2018). Streamlines passing through lesions were isolated to define the affected connectivity.

The disconnectome graph was built using brain regions from the Brainnetome atlas (Fan et al. 2016) as nodes, whilst edges were weighted by the percent of unaffected streamlines connecting two nodes relative to the atlas connectivity. Topological features were extracted from the disconnectome graph and their Spearman’s correlations with TLV and EDSS were computed.

Results

Transitivity (T) and global efficiency (GE) decreased for larger TLV (R=-0.42 and R=-0.78), whereas the average shortest path length (PL) increased (R=0.78). When looking at correlations with EDSS, T (R=-0.17), GE (R=-0.24) and PL (R=0.23) showed stronger associations than lesion count (R=0.14) but were comparable to TLV (R=0.23). All correlations were significant (p<0.001).

Conclusions

We proposed an atlas-based disconnectome model which allowed to study connectivity loss in MS patients without requiring dMRI. Overall, patients showed a lower small-worldness and efficiency for larger TLV and worse disability. These observations were consistent with previous studies on diffusion-based connectomes and open new avenues of research for routine clinical data.