SIFT2: Enabling dense quantitative assessment of brain white matter connectivity using streamlines tractography.

Diffusion MRI streamlines tractography allows for the investigation of the brain white matter pathways non-invasively. However a fundamental limitation of this technology is its non-quantitative nature, i.e. the density of reconstructed connections is not reflective of the density of underlying white matter fibres. As a solution to this problem, we have previously published the "spherical-deconvolution informed filtering of tractograms (SIFT)" method, which determines a subset of the streamlines reconstruction such that the streamlines densities throughout the white matter are as close as possible to fibre densities estimated using the spherical deconvolution diffusion model; this permits the use of streamline count as a valid biological marker of connection density. Particular aspects of its performance may have however limited its uptake in the diffusion MRI research community. Here we present an alternative to this method, entitled SIFT2, which provides a more logically direct and computationally efficient solution to the streamlines connectivity quantification problem: by determining an appropriate cross-sectional area multiplier for each streamline rather than removing streamlines altogether, biologically accurate measures of fibre connectivity are obtained whilst making use of the complete streamlines reconstruction.

Lesion segmentation from multimodal MRI using random forest following ischemic stroke.

Understanding structure-function relationships in the brain after stroke is reliant not only on the accurate anatomical delineation of the focal ischemic lesion, but also on previous infarcts, remote changes and the presence of white matter hyperintensities. The robust definition of primary stroke boundaries and secondary brain lesions will have significant impact on investigation of brain-behavior relationships and lesion volume correlations with clinical measures after stroke. Here we present an automated approach to identify chronic ischemic infarcts in addition to other white matter pathologies, that may be used to aid the development of post-stroke management strategies. Our approach uses Bayesian-Markov Random Field (MRF) classification to segment probable lesion volumes present on fluid attenuated inversion recovery (FLAIR) MRI. Thereafter, a random forest classification of the information from multimodal (T1-weighted, T2-weighted, FLAIR, and apparent diffusion coefficient (ADC)) MRI images and other context-aware features (within the probable lesion areas) was used to extract areas with high likelihood of being classified as lesions. The final segmentation of the lesion was obtained by thresholding the random forest probabilistic maps. The accuracy of the automated lesion delineation method was assessed in a total of 36 patients (24 male, 12 female, mean age: 64.57±14.23yrs) at 3months after stroke onset and compared with manually segmented lesion volumes by an expert. Accuracy assessment of the automated lesion identification method was performed using the commonly used evaluation metrics. The mean sensitivity of segmentation was measured to be 0.53±0.13 with a mean positive predictive value of 0.75±0.18. The mean lesion volume difference was observed to be 32.32%±21.643% with a high Pearson's correlation of r=0.76 (p<0.0001). The lesion overlap accuracy was measured in terms of Dice similarity coefficient with a mean of 0.60±0.12, while the contour accuracy was observed with a mean surface distance of 3.06mm±3.17mm. The results signify that our method was successful in identifying most of the lesion areas in FLAIR with a low false positive rate.

Ophthalmological, cognitive, electrophysiological and MRI assessment of visual processing in preterm children without major neuromotor impairment.

Many studies report chronic deficits in visual processing in children born preterm. We investigated whether functional abnormalities in visual processing exist in children born preterm but without major neuromotor impairment (i.e. cerebral palsy). Twelve such children (< 33 weeks gestation or birthweight < 1000 g) without major neuromotor impairment and 12 born full-term controls were assessed at 8-12 years of age by means of ophthalmological assessment (visual acuity, colour vision, stereopsis, stereoacuity, visual fields, ocular motility, motor fusion), cognitive tests of visual-motor, visual-perceptual and visual-spatial skills and pattern-reversal visual evoked potentials (PR-VEPs). All participants also underwent magnetic resonance imaging (MRI) of the brain and neuromotor assessments. No significant differences were found between the groups on the ophthalmological, visual cognitive, neurological, neuromotor or MRI measures. The P100 component of the PR-VEP showed a significantly shorter latency in the preterm compared with the full-term participants. Whilst this P100 finding suggests that subtle abnormalities may exist at the neurophysiological level, we conclude that visual dysfunction is not systematically associated with preterm birth in the context of normal neurological status.

A software tool to generate simulated white matter structures for the assessment of fibre-tracking algorithms.

The assessment of Diffusion-Weighted MRI (DW-MRI) fibre-tracking algorithms has been limited by the lack of an appropriate 'gold standard'. Practical limitations of alternative methods and physical models have meant that numerical simulations have become the method of choice in practice. However, previous numerical phantoms have consisted of separate fibres embedded in homogeneous backgrounds, which do not capture the true nature of white matter. In this paper we describe a method that is able to randomly generate numerical structures consisting of densely packed bundles of fibres, which are much more representative of human white matter, and simulate the DW-MR images that would arise from them under many imaging conditions. User-defined parameters may be adjusted to produce structures with a range of complexities that spans the levels we would expect to find in vivo. These structures are shown to contain many different features that occur in human white matter and which could confound fibre-tracking algorithms, such as tract kissing and crossing. Furthermore, combinations of such features can be sampled by the random generation of many different structures with consistent levels of complexity. The proposed software provides means for quantitative assessment via direct comparison between tracking results and the exact location of the generated fibres. This should greatly improve our understanding of algorithm performance and therefore prove an important tool for fibre tracking development.