FOD-Net: A Deep Learning Method for Fiber Orientation Distribution Angular Super Resolution

Abstract

Mapping the human connectome using fibre-tracking permits the study of brain connectivity and yields new insights into neuroscience. However, reliable connectome reconstruction using diffusion magnetic resonance imaging (dMRI) data acquired by widely available clinical protocols remains challenging, thus limiting the connectome/tractography clinical applications. Here we develop fibre orientation distribution (FOD) network (FOD-Net), a deep-learning-based framework for FOD angular super-resolution. Our method enhances the angular resolution of FOD images computed from common clinical-quality dMRI data, to obtain FODs with quality comparable to those produced from advanced research scanners. Super-resolved FOD images enable superior tractography and structural connectome reconstruction from clinical protocols. The method was trained and tested with high-quality data from the Human Connectome Project (HCP) and further validated with a local clinical 3.0T scanner. Using this method, we improve the angular resolution of FOD images acquired with typical single-shell low-angular-resolution dMRI data (e.g., 32 directions, b=1000 s/mm2) to approximate the quality of FODs derived from time-consuming, multi-shell high-angular-resolution dMRI research protocols. We also demonstrate tractography improvement, removing spurious connections and bridging missing connections. We further demonstrate that connectomes reconstructed by super-resolved FOD achieve comparable results to those obtained with more advanced dMRI acquisition protocols, on both HCP and clinical 3T data. Advances in deep-learning approaches used in FOD-Net facilitate the generation of high quality tractography/connectome analysis from existing clinical MRI environments.

Building a tissue‐unbiased brain template of fiber orientation distribution and tractography with multimodal registration

“…2,3 For example, both the Human Connectome Project (HCP) 4 and UK Biobank 5 use multishell dMRI as an essential imaging modality to depict the wiring of the human brain. Different from T 1 -weighted (T1w) and T 2 -weighted (T2w) MRI, which mostly capture anatomical details in the cerebral cortex and subcortical nucleus, dMRI provides rich information about the organization 6 in white matter. The dMRI has provided many effective metrics to study whiter matter.…”

Building a Tissue-unbiased Brain Template of Fibre Orientation Distribution and Tractography with Multimodal Registration

“…For example, both the Human Connectome Project (HCP) (Sotiropoulos et al, 2013) and UK Biobank (Alfaro-Almagro et al, 2018) employ multi-shell dMRI as an essential imaging modality to depict the wiring of the human brain. Different from T1-weighted (T1w) and T2- weighted (T2w) MRI, which mostly capture anatomical details in the cerebral cortex and subcortical nucleus, dMRI provides rich information about the organization (Zeng et al, 2022) in white matter. The dMRI has provided many effective metrics to study whiter matter; for example, tensor-based scalar metrics, such as fractional anisotropy (FA) and mean diffusivity (MD)(Alexander et al, 2007), have been widely used for white matter characterization.…”

Angular Super-Resolution in Diffusion MRI with a 3D Recurrent Convolutional Autoencoder

“…This has the advantage of simplifying the task, but limits the ability of the super-resolved data to be used in different analysis techniques. Both Lucena et al (2020) and Zeng et al (2021) used single-shell data and convolutional neural network (CNN) architectures to infer fibre orientation distribution (FOD) data with similar quality to a multi-shell acquisition (Tournier et al, 2007). Similarly, Golkov et al (2016), Chen et al (2020), andYe et al (2020) developed deep architectures to infer metrics from models such as neurite orientation dispersion and density imaging (NODDI) (Zhang et al, 2012) and others, that would otherwise be unavailable with single-shell data.…”