Atlas-powered deep learning (ADL) -- application to diffusion weighted MRI

TL;DR

This paper introduces a novel framework combining deep learning and atlas-based methods to improve biomarker estimation in diffusion weighted MRI, demonstrating superior accuracy and robustness over existing techniques.

Contribution

The paper presents the first integration of deep learning with atlas-based modeling for biomarker estimation in diffusion MRI, enhancing accuracy and robustness.

Findings

Outperforms standard and recent deep learning methods in biomarker estimation.

More robust to measurement down-sampling.

Achieves high accuracy in estimating fractional anisotropy and neurite orientation dispersion.

Abstract

Deep learning has a great potential for estimating biomarkers in diffusion weighted magnetic resonance imaging (dMRI). Atlases, on the other hand, are a unique tool for modeling the spatio-temporal variability of biomarkers. In this paper, we propose the first framework to exploit both deep learning and atlases for biomarker estimation in dMRI. Our framework relies on non-linear diffusion tensor registration to compute biomarker atlases and to estimate atlas reliability maps. We also use nonlinear tensor registration to align the atlas to a subject and to estimate the error of this alignment. We use the biomarker atlas, atlas reliability map, and alignment error map, in addition to the dMRI signal, as inputs to a deep learning model for biomarker estimation. We use our framework to estimate fractional anisotropy and neurite orientation dispersion from down-sampled dMRI data on a test cohort of 70 newborn subjects. Results show that our method significantly outperforms standard estimation methods as well as recent deep learning techniques. Our method is also more robust to stronger measurement down-sampling factors. Our study shows that the advantages of deep learning and atlases can be synergistically combined to achieve unprecedented accuracy in biomarker estimation from dMRI data.