PROBABILISTIC PREDICTION OF NEUROLOGICAL DISORDERS WITH A STATISTICAL ASSESSMENT OF NEUROIMAGING DATA MODALITIES.

For many neurological disorders, prediction of disease state is an important clinical aim. Neuroimaging provides detailed information about brain structure and function from which such predictions may be statistically derived. A multinomial logit model with Gaussian process priors is proposed to: (i) predict disease state based on whole-brain neuroimaging data and (ii) analyze the relative informativeness of different image modalities and brain regions. Advanced Markov chain Monte Carlo methods are employed to perform posterior inference over the model. This paper reports a statistical assessment of multiple neuroimaging modalities applied to the discrimination of three Parkinsonian neurological disorders from one another and healthy controls, showing promising predictive performance of disease states when compared to nonprobabilistic classifiers based on multiple modalities. The statistical analysis also quantifies the relative importance of different neuroimaging measures and brain regions in discriminating between these diseases and suggests that for prediction there is little benefit in acquiring multiple neuroimaging sequences. Finally, the predictive capability of different brain regions is found to be in accordance with the regional pathology of the diseases as reported in the clinical literature.

Differential corticospinal tract degeneration in homozygous 'D90A' SOD-1 ALS and sporadic ALS.

BACKGROUND: The homogeneous genotype and stereotyped phenotype of a unique familial form of amyotrophic lateral sclerosis (ALS) (patients homozygous for aspartate-to-alanine mutations in codon 90 (homD90A) superoxide dismutase 1) provides an ideal model for studying genotype/phenotype interactions and pathological features compared with heterogeneous apparently sporadic ALS. The authors aimed to use diffusion tensor tractography to quantify and compare changes in the intracerebral corticospinal tracts of patients with both forms of ALS, building on previous work using whole-brain voxelwise group analysis. METHOD: 21 sporadic ALS patients, seven homD90A patients and 20 healthy controls underwent 1.5 T diffusion tensor MRI. Patients were assessed using 'upper motor neuron burden,' El Escorial and ALSFR-R scales. The intracranial corticospinal tract was assessed using diffusion tensor tractography measures of fractional anisotropy (FA), mean diffusivity, and radial and axial diffusivity obtained from its entire length. RESULTS: Corticospinal tract FA was reduced in sporadic ALS patients compared with both homD90A ALS patients and controls. The diffusion measures in sporadic ALS patients were consistent with anterograde (Wallerian) degeneration of the corticospinal tracts. In sporadic ALS, corticospinal tract FA was related to clinical measures. Despite a similar degree of clinical upper motor neuron dysfunction and disability in homD90A ALS patients compared with sporadic ALS, there were no abnormalities in corticospinal tract diffusion measures compared with controls. CONCLUSIONS: Diffusion tensor tractography has shown axonal degeneration within the intracerebral portion of the corticospinal tract in sporadic ALS patients, but not those with a homogeneous form of familial ALS. This suggests significant genotypic influences on the phenotype of ALS and may provide clues to slower progression of disease in homD90A patients.

Quantitative magnetic resonance spectroscopic imaging in Parkinson's disease, progressive supranuclear palsy and multiple system atrophy.

BACKGROUND AND PURPOSE: Magnetic resonance spectroscopy (MRS) allows the measurement of a number of brain tissue metabolites in vivo, including N-acetylaspartate (NAA), a putative marker of neuronal integrity. Unlike single voxel MRS, magnetic resonance spectroscopic imaging (MRSI) enables quantification of these metabolites simultaneously from multiple anatomically localized voxels. Both single voxel MRS and MRSI allow the absolute quantification of these metabolites and, when combined with tissue segmentation, can give accurate metabolite concentrations even in the presence of partial volume effects from nearby cerebrospinal fluid. METHODS: Using MRSI with cubic voxels with a nominal volume of 1.0 cm(3), we tested the hypothesis that concentrations of NAA in the basal ganglia in multiple system atrophy (MSA) and progressive supranuclear palsy (PSP) would show differences compared to Parkinson's disease (IPD). NAA values (in mM) from MRSI voxels centred to the putamen, pallidum and thalamus were obtained from 11 patients with IPD, 11 with MSA-P, six with MSA-C, 13 with PSP and 18 controls. The mean concentrations of NAA and its bulk grey and white matter values were also estimated over the whole brain slab. RESULTS: N-acetylaspartate concentrations in the pallidum, putamen and lentiform nucleus were significantly lower in patients with MSA-P and PSP compared to IPD and controls. The putaminal values were also significantly reduced in PSP compared to MSA-P. There were no significant differences between groups in the thalamus and over the whole brain slab. CONCLUSION: Our findings support the notion that MRSI can potentially quantify basal ganglia cellular pathology in MSA and PSP.

Altered cortical activation during a motor task in ALS. Evidence for involvement of central pathways.

OBJECTIVE: To test the hypothesis that patients with amyotrophic lateral sclerosis (ALS) show increased cortical activation during a motor task compared to both healthy controls and patients with muscle weakness due to peripheral lesions. METHODS: Functional magnetic resonance imaging (fMRI) was used to measure activation during a block design paradigm contrasting right hand movements against rest in sixteen patients with ALS, seventeen healthy controls and nine patients with peripheral lesions. The groups were matched for age and gender and the two patient groups were matched for their degree of upper limb weakness. Analysis used a non-parametric approach to perform a 3 way hypothesis-driven comparison between the groups. RESULTS: During the motor task, patients with ALS showed increased cortical activation bilaterally, extending from the sensorimotor cortex [Brodmann areas (BA) 1, 2, 4] posteriorly into the inferior parietal lobule (BA 40) and inferiorly to the superior temporal gyrus (BA 22) when compared to peripheral lesion patients and controls. In addition, ALS patients showed reduced activation in the dorsolateral prefrontal cortex (DLPFC) extending to anterior and medial frontal cortex (BA 8, 9, 10, 32). CONCLUSIONS: We conclude that alterations in cortical function in ALS differ in sensorimotor and prefrontal regions. Importantly, we have shown that these changes do not reflect confounding by weakness or task difficulty, but are likely to be related to upper motor neuron pathology in ALS.

Measuring brain stem and cerebellar damage in parkinsonian syndromes using diffusion tensor MRI.

OBJECTIVE: To use diffusion tensor MRI to quantify and compare degeneration of the pons and cerebellar peduncles in multiple system atrophy (MSA), progressive supranuclear palsy (PSP), and Parkinson disease (PD) and to relate changes in diffusion measures to clinical features and localized atrophy. METHODS: We used a region-of-interest approach to measure changes in fractional anisotropy and mean diffusivity in the middle cerebellar peduncles, decussation of the superior cerebellar peduncles, and pons in 17 patients with MSA, 17 with PSP, 12 with PD, and 12 healthy volunteers. We also evaluated atrophy of the cerebellar peduncles and pons on T2-weighted magnetic resonance images in patients with MSA and PSP. RESULTS: In MSA, fractional anisotropy was markedly reduced in the middle cerebellar peduncles, and mean diffusivity increased both here and in the pons compared with other groups, whereas in PSP, mean diffusivity was strikingly increased in the decussation of superior cerebellar peduncles. Cerebellar ataxia was related to mean diffusivity in the middle cerebellar peduncles (r = 0.71, p = 0.001) and pons (r = 0.60, p = 0.01) in MSA. Diffusion measures were related to localized atrophy in both MSA and PSP. CONCLUSIONS: Diffusion tensor MRI can be used to quantify neurodegenerative processes in different brain stem and cerebellar structures in multiple system atrophy and progressive supranuclear palsy during life, and may have diagnostic value. Larger studies of early, undifferentiated parkinsonian syndromes are indicated to provide estimates of the relative diagnostic value of diffusion measures, atrophy measures, and visual assessment of scans.