Structure supports function: informing directed and dynamic functional connectivity with anatomical priors

The dynamic repertoire of functional brain networks is constrained by the underlying topology of structural connections. Despite this intrinsic relationship between structural (SC) and functional connectivity (FC), integrative and multimodal approaches to combine the two remain limited. Here, we propose a new adaptive filter for estimating dynamic and directed FC using structural connectivity information as priors. We tested the filter in rat epicranial recordings and human event-related EEG data, using SC priors from a meta-analysis of tracer studies and diffusion tensor imaging metrics, respectively. We show that, particularly under conditions of low signal-to-noise ratio, SC priors can help to refine estimates of directed FC, promoting sparse functional networks that combine information from structure and function. In addition, the proposed filter provides intrinsic protection against SC-related false negatives, as well as robustness against false positives, representing a valuable new tool for multimodal imaging in the context of dynamic and directed FC analysis.

Fast Fiber Orientation Estimation in Diffusion MRI from kq-Space Sampling and Anatomical Priors

High spatio-angular resolution diffusion MRI (dMRI) has been shown to provide accurate identification of complex neuronal fiber configurations, albeit, at the cost of long acquisition times. We propose a method to recover intra-voxel fiber configurations at high spatio-angular resolution relying on a 3D kq-space under-sampling scheme to enable accelerated acquisitions. The inverse problem for the reconstruction of the fiber orientation distribution (FOD) is regularized by a structured sparsity prior promoting simultaneously voxel-wise sparsity and spatial smoothness of fiber orientation. Prior knowledge of the spatial distribution of white matter, gray matter, and cerebrospinal fluid is also leveraged. A minimization problem is formulated and solved via a stochastic forward–backward algorithm. Simulations and real data analysis suggest that accurate FOD mapping can be achieved from severe kq-space under-sampling regimes potentially enabling high spatio-angular resolution dMRI in the clinical setting.

Structure supports function: informing directed and dynamic functional connectivity with anatomical priors

The dynamic repertoire of functional brain networks is constrained by the underlying topology of structural connections: the lack of a direct structural link between two brain regions prevents direct functional interactions. Despite the intrinsic relationship between structural (SC) and functional connectivity (FC), integrative and multimodal approaches to combine the two remain limited, especially for electrophysiological data. In the present work, we propose a new linear adaptive filter for estimating dynamic and directed FC using structural connectivity information as priors. We tested the filter in rat epicranial recordings and human event-related EEG data, using SC priors from a meta-analysis of tracer studies and diffusion tensor imaging metrics, respectively. Our results show that SC priors increase the resilience of FC estimates to noise perturbation while promoting sparser networks under biologically plausible constraints. The proposed filter provides intrinsic protection against SC-related false negatives, as well as robustness against false positives, representing a valuable new method for multimodal imaging and dynamic FC analysis.

Exploiting Global Structure Information to Improve Medical Image Segmentation

In this paper, we propose a method to enhance the performance of segmentation models for medical images. The method is based on convolutional neural networks that learn the global structure information, which corresponds to anatomical structures in medical images. Specifically, the proposed method is designed to learn the global boundary structures via an autoencoder and constrain a segmentation network through a loss function. In this manner, the segmentation model performs the prediction in the learned anatomical feature space. Unlike previous studies that considered anatomical priors by using a pre-trained autoencoder to train segmentation networks, we propose a single-stage approach in which the segmentation network and autoencoder are jointly learned. To verify the effectiveness of the proposed method, the segmentation performance is evaluated in terms of both the overlap and distance metrics on the lung area and spinal cord segmentation tasks. The experimental results demonstrate that the proposed method can enhance not only the segmentation performance but also the robustness against domain shifts.

Anatomically informed multi-level fiber tractography

Diffusion weighted MR imaging can assist preoperative planning by reconstructing the trajectory of eloquent fiber pathways. A common task is the delineation of the corticospinal tract in its full extent because lesions to this bundle can severely affect the quality of life. However, this is challenging as existing tractography algorithms typically produce either incomplete results or multiple false-positive tracts. In this work, we suggest a novel approach to fiber tractography that reconstructs multi-level structures by progressively taking into account previously unused fiber orientations. Anatomical priors are used in order to minimize the number of false-positive pathways. The devised method was evaluated on synthetic data with different noise levels. Additionally, it was tested on in-vivo data by reconstructing the corticospinal tract and it was compared to conventional deterministic and probabilistic approaches. The corticospinal tract reconstructed by our method includes lateral projections that could not be observed with deterministic methods, while avoiding spurious tracts reconstructed by probabilistic tractography. Furthermore, the proposed algorithm preserves the neuroanatomical topology of the pathways to a larger extent as compared to probabilistic tractography.

Track-To-Learn: A general framework for tractography with deep reinforcement learning

Diffusion MRI tractography is currently the only non-invasive tool able to assess the white-matter structural connectivity of a brain. Since its inception, it has been widely documented that tractography is prone to producing erroneous tracks while missing true positive connections. Anatomical priors have been conceived and implemented in classical algorithms to try and tackle these issues, yet problems still remain and the conception and validation of these priors is very challenging. Recently, supervised learning algorithms have been proposed to learn the tracking procedure implicitly from data, without relying on anatomical priors. However, these methods rely on labelled data that is very hard to obtain. To remove the need for such data but still leverage the expressiveness of neural networks, we introduce Track-To-Learn: A general framework to pose tractography as a deep reinforcement learning problem. Deep reinforcement learning is a type of machine learning that does not depend on ground-truth data but rather on the concept of “reward”. We implement and train algorithms to maximize returns from a reward function based on the alignment of streamlines with principal directions extracted from diffusion data. We show that competitive results can be obtained on known data and that the algorithms are able to generalize far better to new, unseen data, than prior machine learning-based tractography algorithms. To the best of our knowledge, this is the first successful use of deep reinforcement learning for tractography.