q-Space Deep Learning for Twelve-Fold Shorter and Model-Free Diffusion MRI Scans

Abstract

Abstract. Diffusion MRI uses a multi-step data processing pipeline. With certain steps being prone to instabilities, the pipeline relies on considerable amounts of partly redundant input data, which requires long acquisition time. This leads to high scan costs and makes advanced diffusion models such as diffusion kurtosis imaging (DKI) and neurite orientation dispersion and density imaging (NODDI) inapplicable for children and adults who are uncooperative, uncomfortable or unwell. We demonstrate how deep learning, a group of algorithms in the field of artificial neural networks, can be applied to reduce diffusion MRI data processing to a single optimized step. This method allows obtaining scalar measures from advanced models at twelve-fold reduced scan time and detecting abnormalities without using diffusion models.

Deep Learning in Neuroradiology

“…A deep learning approach can reduce imaging duration 12-fold by predicting final parameter maps (fractional anisotropy, mean diffusivity, and so forth) from relatively few angular directions. 17 By acquiring paired arterial spin-labeling (ASL) CBF images with 2 and 30 minutes of acquisition time, our group has trained a deep network to boost the SNR of ASL significantly (Fig 5).…”

Guest Editorial Deep Learning in Medical Imaging: Overview and Future Promise of an Exciting New Technique

“…The relationship between the diffusion-weighted signal and microstructural tissue properties is non-trivial. Golkov et al [35] demonstrate that with the use of a DNN such relationships may in fact be revealed: DWIs are directly used as inputs rather than using scalar measures obtained from model fitting. The work shows microstructure prediction on a voxel-by-voxel basis as well as automated model-free segmentation from DWI values, into healthy tissues and MS lesions.…”

“…In an exploratory work, Golkov et al [35] provide an initial proof-of-concept, applying DL to reduce diffusion MRI data processing to a single optimized step. They show that this modification enables one to obtain scalar measures from advanced models at twelve-fold reduced scan time and to detect abnormalities without using diffusion models.…”

Fast imaging of mean, axial and radial diffusion kurtosis

“…We recently proposed a fast imaging framework, capable of estimating mean kurtosis on the basis of only 13 (139 protocol (Hansen et al, 2013a)) or 19 (199 protocol (Hansen et al, 2015)) diffusion weighted images. In contrast to other methods suggested to accelerate DKI by reducing the data requirements (Giannelli and Toschi, 2016; Golkov et al, 2015; Tachibana et al, 2015), our approach does not rely on additional assumptions or approximations beyond the few already inherent to DKI. Our approach was later adopted and validated in a rat model of stroke (Sun et al, 2014), and applied in human cancer patients, where it was shown to discriminate among different brain tumor types (Tietze et al, 2015).…”