A Genetic Algorithm for the selection of structural MRI features for classification of Mild Cognitive Impairment and Alzheimer's Disease

This work investigates the problem of feature selection in neuroimaging features from structural MRI brain images for the classification of subjects as healthy controls, suffering from Mild Cognitive Impairment or Alzheimer’s Disease. A Genetic Algorithm wrapper method for feature selection is adopted in conjunction with a Support Vector Machine classifier. In very large feature sets, feature selection is found to be redundant as the accuracy is often worsened when compared to an Support Vector Machine with no feature selection. However, when just the hippocampal subfields are used, feature selection shows a significant improvement of the classification accuracy. Three-class Support Vector Machines and two-class Support Vector Machines combined with weighted voting are also compared with the former and found more useful. The highest accuracy achieved at classifying the test data was 65.5% using a genetic algorithm for feature selection with a three-class Support Vector Machine classifier.

Structural features of condensed tannins affect in vitro ruminal methane production and fermentation characteristics

An in vitro study was conducted to investigate the effects of condensed tannins (CT) structural properties, i.e. average polymer size (or mean degree of polymerization); percentage of cis flavan-3-ols and percentage of prodelphinidins in CT extracts on methane production (CH4) and fermentation characteristics. CT were extracted from eight plants in order to obtain different CT types: black currant leaves, goat willow leaves, goat willow twigs, pine bark, red currant leaves, sainfoin plants, weeping willow catkins and white clover flowers. They were analysed for CT content and CT composition by thiolytic degradation, followed by HPLC analysis. Grass silage was used as a control substrate. Condensed tannins were added to the substrate at a concentration of 40 g/kg, with or without polyethylene glycol (+ or −PEG 6000 treatment) to inactivate tannins, and then incubated for 72 h in mixed buffered rumen fluid from three different lactating dairy cows per run. Total cumulative gas production (GP) was measured by an automated gas production system. During the incubation, 12 gas samples (10 μl) were collected from each bottle headspace at 0, 2, 4, 6, 8, 12, 24, 30, 36, 48, 56 and 72 h of incubation and analyzed for CH4. A modified Michaelis–Menten model was fitted to the CH4 concentration patterns and model estimates were used to calculate total cumulative CH4 production (GPCH4). Total cumulative gas production and GPCH4 curves were fitted using biphasic and monophasic modified Michaelis-Menten models, respectively. Addition of PEG increased GP, GPCH4, and CH4 concentration compared to the −PEG treatment. All CT types reduced GPCH4 and CH4 concentration. All CT increased the half time of GP and GPCH4. Moreover, all CT decreased the maximum rate of fermentation for GPCH4 and rate of substrate degradation. The correlation between CT structure and GPCH4 and fermentation characteristics showed that the proportion of prodelphinidins within CT had the largest effect on fermentation characteristics, followed by average 27 polymer size and percentage of cis-flavan-3-ols.