Validating atlas-based lesion disconnectomics in multiple sclerosis: A retrospective multi-centric study

Abstract

The translational potential of MR-based connectivity modelling is limited by the need for advanced diffusion imaging, which is not part of clinical protocols for many diseases. In addition, where diffusion data is available, brain connectivity analyses rely on tractography algorithms which imply two major limitations. First, tracking algorithms are known to be sensitive to the presence of white matter lesions and therefore leading to interpretation pitfalls and poor inter-subject comparability in clinical applications such as multiple sclerosis. Second, tractography quality is highly dependent on the acquisition parameters of diffusion sequences, leading to a trade-off between acquisition time and tractography precision. Here, we propose an atlas-based approach to study the interplay between structural disconnectivity and lesions without requiring individual diffusion imaging. In a multi-centric setting involving three distinct multiple sclerosis datasets (containing both 1.5 T and 3 T data), we compare our atlas-based structural disconnectome computation pipeline to disconnectomes extracted from individual tractography and explore its clinical utility for reducing the gap between radiological findings and clinical symptoms in multiple sclerosis. Results using topological graph properties showed that overall, our atlas-based disconnectomes were suitable approximations of individual disconnectomes from diffusion imaging. Small-worldness was found to decrease for larger total lesion volumes thereby suggesting a loss of efficiency in brain connectivity of MS patients. Finally, the global efficiency of the created brain graph, combined with total lesion volume, allowed to stratify patients into subgroups with different clinical scores in all three cohorts.

Trial registration: ClinicalTrials.gov NCT01592474.

Keywords: Brain graphs; Diffusion imaging; Disconnectome; Network neuroscience; Structural connectivity; Topology.

Copyright © 2021 The Author(s). Published by Elsevier Inc. All rights reserved.

Figures

Graphical abstract

Fig. 1

Extraction and modelling of a disconnectome. A. Simplified representation of the atlas-based tractogram connectivity of three brain regions i, j and k and their respective streamlines. At is the associated adjacency matrix of the atlas connectome, where each element At(i,j) represents the number of streamlines connecting the pair of regions (i,j). B. Overlapping of the atlas tractogram with the lesion mask. Streamlines passing through the lesion L are highlighted in red. C. Affected streamlines are isolated. Ac is the adjacency matrix of the affected streamlines, where each element Ac(I,j) represents the number of affected streamlines connecting the pair of regions (i,j) D. The brain graph representation where brain areas i, j and k are represented respectively by nodes Vi Vj and Vk and edges are weighted by the relative number of affected streamlines. The adjacency matrix of the remaining connectivity is defined as the relative difference between At and Ac. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Extraction and representation of a disconnectome. A. The streamlines intersecting the lesion mask are isolated from the healthy tractography atlas. B. Affected streamlines are overlaid with an anatomical atlas to create a brain graph representation of disconnectivity. C. Lobe-wise summary representation of disconnectivity, visualized using Circos (Connors et al., 2009).

Fig. 3

Flow diagram showing the applicability conditions of the atlas-based disconnectome model, based on age, brain atrophy and brain segmentation quality estimated from MorphoBox (Schmitter et al., 2015). f(x) is the 3% percentile of normative ranges. Ni: initial number of patients. N0: number of patients discarded due to poor registration quality. N1: number of patients discarded due to bad segmentation quality. N2: number of patients younger than the age range covered by the tractography atlas, who are discarded due to abnormally low brain volume. N3: number of patients inside the atlas age range discarded due to abnormally low brain volume. N4: number of patients older than the age range covered by the tractography atlas, who are discarded due to abnormally low brain volume.

Fig. 4

Large scale topological features extracted from atlas-based disconnectome graph (y-axis) plotted against equivalent metrics extracted from individual tractography-based disconnectome (x-axis) for patients of the diffusion cohort. Each point is a patient. Disconnectome features derived from manual lesion segmentation are shown in yellow, and automated lesion segmentation with LeManPV in blue. ICC(3,1) and Spearman’s correlation are reported with their respective significance level after correction for multiple comparison. * p 

Fig. 5

Variation of correlation of small-scale features extracted from individual versus atlas-based tractography with lesion load for the diffusion dataset. Each point is a patient. Spearman correlation and ICC(3,1) agreement between the two methods are computed across nodes for all patients and are shown on the y-axis. The total lesion volume (TLV) estimated from either manual (in yellow) or LeMan-PV lesion segmentations (in blue) on the x-axis. Labels in the graph indicate patient with generally high or low agreement, for which details are provided in Supplementary Fig. 3.

Fig. 6

Variation of large scale topological features (y-axis) with total lesion volume (TLV, x-axis) for patients in all datasets (diffusion cohort in grey, 1.5 T cohort in blue and 3 T cohort in yellow). Spearman’s R and p-value are reported for all datasets. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 7

Variation of large scale topological features (y-axis) with Expanded Disability Status Scale (EDSS, x-axis) for patients in all datasets (diffusion cohort in grey, 1.5 T cohort in blue and 3 T cohort in yellow). Spearman’s R and p-value are reported for all datasets. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 8

EDSS distribution in all cohorts when stratified according to total lesion volume (TLV) and the global efficiency (GE) of their remaining connectivity. The cut-offs used to classify patients were the average TLV and GE within each cohort. These cut-offs are reported in Supplementary Table 3.

References

An official website of the United States government

Here's how you know

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Show account info Close Account Logged in as:
username Server error

Sorry, an error has occurred. Please try again in a moment.

Go to .

NCBI Literature Resources

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

Follow NCBI

National Library of Medicine

var nwds_version = "1.2.0"; var meta_nwds_ver = document.createElement('meta'); meta_nwds_ver.name = 'ncbi_nwds_ver'; meta_nwds_ver.content = nwds_version; document.getElementsByTagName('head')[0].appendChild(meta_nwds_ver); var meta_nwds = document.createElement('meta'); meta_nwds.name = 'ncbi_nwds'; meta_nwds.content = 'yes'; document.getElementsByTagName('head')[0].appendChild(meta_nwds); var alertsUrl = "/core/alerts/alerts.js"; if (typeof ncbiBaseUrl !== 'undefined') { alertsUrl = ncbiBaseUrl + alertsUrl; } ncbi.awesome.basePage.init({ userInfo: { isLoggedIn: false, username: "", loginUrl: "https://account.ncbi.nlm.nih.gov/?back_url=https%3A%2F%2Fpubmed.ncbi.nlm.nih.gov%2F34500427%2Freferences%2F%3Ftimestamp%3D1680585033753", logoutUrl: "https://www.ncbi.nlm.nih.gov/account/signout/?back_url=https%3A//pubmed.ncbi.nlm.nih.gov/34500427/references/%3Ftimestamp%3D1680585033753" }, baseUrl: 'https://www.ncbi.nlm.nih.gov', noSessionCookieName: "pm-ns", searchIdCookieName: "pm-sid", adjNavSearchIdCookieName: "pm-adjnav-sid", searchFormSubmittedCookieName: "pm-sfs", isOnSearchPageCookieName: "pm-iosp", backToSearchCookieName: "pm-btsl", clinicalSeeAllCookieName: "pm-csa", jsCookieDomain: "pubmed.ncbi.nlm.nih.gov", userShareRequestFailed: "", noSharedSettings: "myncbi_no_shared_settings", noSharedAccount: "myncbi_no_shared_account" }); jQuery.getScript("https://www.ncbi.nlm.nih.gov/core/alerts/alerts.js", function () { galert(['body > *:last-child']); ncbi.awesome.dismissableNCBIAlert(); }).fail(function() { });