Directed functional connectivity using dynamic graphical models.

There are a growing number of neuroimaging methods that model spatio-temporal patterns of brain activity to allow more meaningful characterizations of brain networks. This paper proposes dynamic graphical models (DGMs) for dynamic, directed functional connectivity. DGMs are a multivariate graphical model with time-varying coefficients that describe instantaneous directed relationships between nodes. A further benefit of DGMs is that networks may contain loops and that large networks can be estimated. We use network simulations and human resting-state fMRI (N = 500) to investigate the validity and reliability of the estimated networks. We simulate systematic lags of the hemodynamic response at different brain regions to investigate how these lags potentially bias directionality estimates. In the presence of such lag confounds (0.4-0.8 s offset between connected nodes), our method has a sensitivity of 72%-77% to detect the true direction. Stronger lag confounds have reduced sensitivity, but do not increase false positives (i.e., directionality estimates of the opposite direction). In human resting-state fMRI, the default mode network has consistent influence on the cerebellar, the limbic and the auditory/temporal networks. We also show a consistent reciprocal relationship between the visual medial and visual lateral network. Finally, we apply the method in a small mouse fMRI sample and discover a highly plausible relationship between areas in the hippocampus feeding into the cingulate cortex. We provide a computationally efficient implementation of DGM as a free software package for R.

Application of a novel tool for diagnosing bile acid diarrhoea.

Bile acid diarrhoea (BAD) is a common disease that requires expensive imaging to diagnose. We have tested the efficacy of a new method to identify BAD, based on the detection of differences in volatile organic compounds (VOC) in urine headspace of BAD vs. ulcerative colitis and healthy controls. A total of 110 patients were recruited; 23 with BAD, 42 with ulcerative colitis (UC) and 45 controls. Patients with BAD also received standard imaging (Se75HCAT) for confirmation. Urine samples were collected and the headspace analysed using an AlphaMOS Fox 4000 electronic nose in combination with an Owlstone Lonestar Field Asymmetric Ion Mobility Spectrometer (FAIMS). A subset was also tested by gas chromatography, mass spectrometry (GCMS). Linear Discriminant Analysis (LDA) was used to explore both the electronic nose and FAIMS data. LDA showed statistical differences between the groups, with reclassification success rates (using an n-1 approach) at typically 83%. GCMS experiments confirmed these results and showed that patients with BAD had two chemical compounds, 2-propanol and acetamide, that were either not present or were in much reduced quantities in the ulcerative colitis and control samples. We believe that this work may lead to a new tool to diagnose BAD, which is cheaper, quicker and easier that current methods.

Differentiating coeliac disease from irritable bowel syndrome by urinary volatile organic compound analysis--a pilot study.

Coeliac disease (CD), a T-cell-mediated gluten sensitive enteropathy, affects ∼ 1% of the UK population and can present with wide ranging clinical features, often being mistaken for Irritable Bowel Syndrome (IBS). Heightened clinical awareness and serological screening identifies those with potential coeliac disease; the diagnosis is confirmed with duodenal biopsies, and symptom improvement with a gluten-free diet. Limitations to diagnosis are false negative serology and reluctance to undergo biopsy. The gut microbiome is altered in several gastrointestinal disorders, causing altered gut fermentation patterns recognisable by volatile organic compounds (VOC) analysis in urine, breath and faeces. We aimed to determine if CD alters the urinary VOC pattern, distinguishing it from IBS. 47 patients were recruited, 27 with established CD, on gluten free diets, and 20 with diarrhoea-predominant IBS (D-IBS). Collected urine was stored frozen in 10 ml aliquots. For assay, the specimens were heated to 40 ± 0.1°C and the headspace analysed by Field Asymmetric Ion Mobility Spectrometry (FAIMS). Machine learning algorithms were used for statistical evaluation. Samples were also analysed using Gas chromatography and mass spectroscopy (GC-MS). Sparse logistic regression showed that FAIMS distinguishes VOCs in CD vs D-IBS with ROC curve AUC of 0.91 (0.83-0.99), sensitivity and specificity of 85% respectively. GCMS showed a unique peak at 4'67 found only in CD, not D-IBS, which correlated with the compound 1,3,5,7 cyclooctatetraene. This study suggests that FAIMS offers a novel, non-invasive approach to identify those with possible CD, and distinguishes from D-IBS. It offers the potential for monitoring compliance with a gluten-free diet at home. The presence of cyclooctatetraene in CD specimens will need further validation.