CoChat
Sign Up
Cited by 138Advanced Neuroimaging Techniques and ApplicationsFunctional Brain Connectivity StudiesAdvanced MRI Techniques and Applications

The DOI matches its Crossref record; 51 of the 51 references it cites were found in a registry or an index.

Microstructure Informed Tractography: Pitfalls and Open Challenges

Open access
Daducci, Alessandro, Dal Palú, Alessandro, Descoteaux, Maxime, Thiran, Jean-Philippe·2016·Frontiers in Neuroscience
DOI 10.3389/fnins.2016.00247PMC PMC4893481
Open this paper in CoChat — free accountOpen PDF

Abstract

One of the major limitations of diffusion MRI tractography is that the fiber tracts recovered by existing algorithms are not truly quantitative. Local techniques for estimating more quantitative features of the tissue microstructure exist, but their combination with tractography has always been considered intractable. Recent advances in local and global modeling made it possible to fill this gap and a number of promising techniques for microstructure informed tractography have been suggested, opening new and exciting perspectives for the quantification of brain connectivity. The ease-of-use of the proposed solutions made it very attractive for researchers to include such advanced methods in their analyses; however, this apparent simplicity should not hide some critical open questions raised by the complexity of these very high-dimensional problems, otherwise some fundamental issues may be pushed into the background. The aim of this article is to raise awareness in the diffusion MRI community, notably researchers working on brain connectivity, about some potential pitfalls and modeling choices that make the interpretation of the outcomes from these novel techniques rather cumbersome. Through a series of experiments on synthetic and real data, we illustrate practical situations where erroneous and severely biased conclusions may be drawn about the connectivity if these pitfalls are overlooked, like the presence of partial/missing/duplicate fibers or the critical importance of the diffusion model adopted. Microstructure informed tractography is a young but very promising technology, and by acknowledging its current limitations as done in this paper, we hope our observations will trigger further research in this direction and new ideas for truly quantitative and biologically meaningful analyses of the connectivity.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.3389/fnins.2016.00247

Cites

51 of 51 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Plumbing metaphor picturing microstructure informed tractography as a tool to evaluate the health condition of the water supply network of a big city.
Figure 1 Plumbing metaphor picturing microstructure informed tractography as a tool to evaluate the health condition of the water supply network of a big city.CC BY · PubMed Central
Figure 2 Parallel between local and global reconstruction. Local reconstruction methods usually recover the fiber configuration in a voxel by expressing the quantity of interest, e.g., ODF, as a linear combination of a given set of response functions (shown here in 2D for simplicity). Daducci et al. (2013, 2014a) showed that also microstructure informed tractography can be expressed using the same formulation where, instead of a single voxel, the dictionary models the whole dMRI image as a superimposition of the signal arising from all the fibers in a tractogram.
Figure 2 Parallel between local and global reconstruction. Local reconstruction methods usually recover the fiber configuration in a voxel by expressing the quantity of interest, e.g., ODF, as a linear combination of a given set of response functions (shown here in 2D for simplicity). Daducci et al. (2013, 2014a) showed that also microstructure informed tractography can be expressed using the same formulation where, instead of a single voxel, the dictionary models the whole dMRI image as a superimposition of the signal arising from all the fibers in a tractogram.CC BY · PubMed Central
Figure 3 Improper dictionary: missing atoms. (A) Schematic illustration of a dictionary with a missing atom (number 3 in Figure 2). Estimated coefficients (B) and ODF (C) when using it to reconstruct the voxel configuration in Figure 2. (D) Parallel with global reconstruction: the common difficulty of classical tractography algorithms in tracking the callosal projection fibers (yellow boxes) is a typical scenario that leads to missing atoms (i.e., fibers) in the dictionary.
Figure 3 Improper dictionary: missing atoms. (A) Schematic illustration of a dictionary with a missing atom (number 3 in Figure 2). Estimated coefficients (B) and ODF (C) when using it to reconstruct the voxel configuration in Figure 2. (D) Parallel with global reconstruction: the common difficulty of classical tractography algorithms in tracking the callosal projection fibers (yellow boxes) is a typical scenario that leads to missing atoms (i.e., fibers) in the dictionary.CC BY · PubMed Central
Figure 4 Improper dictionary: duplicate atoms. (A) Dictionary with replicas of an atom (number 3 in Figure 2). Estimated coefficients (B) and ODF (C) when using it to reconstruct the voxel configuration in Figure 2. (D) Parallel with global reconstruction: the trajectories of an easy-to-track bundle (yellow line) can be reconstructed many times in a tractogram and the corresponding atoms can be assigned very small contributions; this could lead to conclude that each duplicate fiber is spurious.
Figure 4 Improper dictionary: duplicate atoms. (A) Dictionary with replicas of an atom (number 3 in Figure 2). Estimated coefficients (B) and ODF (C) when using it to reconstruct the voxel configuration in Figure 2. (D) Parallel with global reconstruction: the trajectories of an easy-to-track bundle (yellow line) can be reconstructed many times in a tractogram and the corresponding atoms can be assigned very small contributions; this could lead to conclude that each duplicate fiber is spurious.CC BY · PubMed Central
Figure 5 Toy problem to illustrate some issues of partial fibers on connectivity analysis. Tracts connecting the four gray-matter regions R1, …, R4 are marked as solid lines and those stopping prematurely in the white matter as dashed. It is well-known that the (random) number of partial fibers recovered with tractography can significantly bias the estimation; unfortunately, if these partial fibers are not properly handled, also microstructure informed tractography techniques can result ineffective.
Figure 5 Toy problem to illustrate some issues of partial fibers on connectivity analysis. Tracts connecting the four gray-matter regions R1, …, R4 are marked as solid lines and those stopping prematurely in the white matter as dashed. It is well-known that the (random) number of partial fibers recovered with tractography can significantly bias the estimation; unfortunately, if these partial fibers are not properly handled, also microstructure informed tractography techniques can result ineffective.CC BY · PubMed Central
Figure 6 Importance of the forward-model. The toy problem in (A) is used to analyze the behavior of two different forward-models. The model in (B) assumes that the signal in each voxel originates exclusively from the tracts of the two fascicles F1 and F2 passing through it, whereas (C) considers all the possible water pools that can contribute to the dMRI signal, in particular the extra-cellular space (EC) around the axons. (D96F) Analogous situation in the brain using the HCP datasets. See description in the text for details.
Figure 6 Importance of the forward-model. The toy problem in (A) is used to analyze the behavior of two different forward-models. The model in (B) assumes that the signal in each voxel originates exclusively from the tracts of the two fascicles F1 and F2 passing through it, whereas (C) considers all the possible water pools that can contribute to the dMRI signal, in particular the extra-cellular space (EC) around the axons. (D96F) Analogous situation in the brain using the HCP datasets. See description in the text for details.CC BY · PubMed Central
Figure 7 Biological plausibility. The fitting error (NRMSE) provides us with useful information about how well a tractogram explains the measured dMRI signal, but nothing concerning the biological plausibility of the estimated tract contributions. To this end, it can be very useful to inspect, for example, the fractions of the intra- (icvf) and extra-cellular (ecvf) compartments in each voxel as predicted by global optimization and compare to the known anatomy. For example, blue arrows highlight areas where the estimated fractions follow the expected pattern of neuronal density (see text for details), whereas white arrows point to the same regions where these contributions appear unrealistic in the last column.
Figure 7 Biological plausibility. The fitting error (NRMSE) provides us with useful information about how well a tractogram explains the measured dMRI signal, but nothing concerning the biological plausibility of the estimated tract contributions. To this end, it can be very useful to inspect, for example, the fractions of the intra- (icvf) and extra-cellular (ecvf) compartments in each voxel as predicted by global optimization and compare to the known anatomy. For example, blue arrows highlight areas where the estimated fractions follow the expected pattern of neuronal density (see text for details), whereas white arrows point to the same regions where these contributions appear unrealistic in the last column.CC BY · PubMed Central
Figure 8 Impact of signal inhomogeneities on global analyses. In this example, the artificial modulation of the intensities, e.g., along the arrow direction, leads to a spatially-varying scaling of the estimated parameter maps that might be erroneously interpreted as a real change in the underlying connectivity.
Figure 8 Impact of signal inhomogeneities on global analyses. In this example, the artificial modulation of the intensities, e.g., along the arrow direction, leads to a spatially-varying scaling of the estimated parameter maps that might be erroneously interpreted as a real change in the underlying connectivity.CC BY · PubMed Central
Figure 9 Fiber length and convergence. (A) The fibers in a tractogram vary greatly in length and thus their relative contributions to the dMRI image are quite different. (B) This imbalance may lead to very slow convergence and to a premature termination of the optimization process.
Figure 9 Fiber length and convergence. (A) The fibers in a tractogram vary greatly in length and thus their relative contributions to the dMRI image are quite different. (B) This imbalance may lead to very slow convergence and to a premature termination of the optimization process.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 51 references

Cited by

Sorted by citations

See all 138

Related

Metadata

Authors
Daducci, A., Dal Palú, A., Descoteaux, M., & Thiran, J.-P.
Affiliations
  • Université de Sherbrooke (CA)
  • Centre Hospitalier Universitaire Vaudois (CH)
Corresponding author
Daducci, Alessandro
ORCID
Venue
Frontiers in Neuroscience
Publisher
Frontiers Media SA
Published
2016-06-06
Volume
10
Pages
247-247
ISSN
1662-453X (ISSN-L)
DOI
10.3389/fnins.2016.00247
PMC
PMC4893481
OpenAlex
W2404527239
Cited by
138
References
51
Citations by year
2026: 9 · 2025: 17 · 2024: 9 · 2023: 8 · 2022: 8 · 2021: 12 · 2020: 14 · 2019: 23 · 2018: 16 · 2017: 21 · 2016: 1
FWCI
17.65
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • Frontiers in Neuroscience (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • Figshare (submittedVersion, cc-by)
  • Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion)
  • Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion, cc-by-nc-nd)
  • Figshare (cc-by)
Funders
  • École Polytechnique Fédérale de Lausanne
  • Centre d'Imagerie BioMédicale
Topics
  • Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Advanced Neuroimaging Techniques and Applications
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Functional Brain Connectivity Studies
  • Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Advanced MRI Techniques and Applications
Keywords
diffusion MRI, brain connectivity

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review