Search

The coronavirus disease (COVID-19) is an infectious disease caused by the SARS-CoV-2 virus. COVID-19 is found to be the most infectious disease in last few decades. This disease has infected millions of people worldwide. The inadequate availability and the limited sensitivity of the testing kits have motivated the clinicians and the scientist to use Computer Tomography (CT) scans to screen COVID-19. Recent advances in technology and the availability of deep learning approaches have proved to be very promising in detecting COVID-19 with increased accuracy. However, deep learning approaches require a huge labeled training dataset, and the current availability of benchmark COVID-19 data is still small. For the limited training data scenario, the CNN usually overfits after several iterations. Hence, in this work, we have investigated different pre-trained network architectures with transfer learning for COVID-19 detection that can work even on a small medical imaging dataset. Various variants of the pre-trained ResNet model, namely ResNet18, ResNet50, and ResNet101, are investigated in the current paper for the detection of COVID-19. The experimental results reveal that transfer learned ResNet50 model outperformed other models by achieving a recall of 98.80% and an F1-score of 98.41%. To further improvise the results, the activations from different layers of best performing model are also explored for the detection using the support vector machine, logistic regression and K-nearest neighbor classifiers. Moreover, a classifier fusion strategy is also proposed that fuses the predictions from the different classifiers via majority voting. Experimental results reveal that via using learned image features and classification fusion strategy, the recall, and F1-score have improvised to 99.20% and 99.40%.

Intestinal flora of wild chimpanzee has not been studied. Fecal flora analyses currently give insight to this environment. We collected feces from twelve human-habituated wild chimpanzees in each of three age groups: four juveniles, four sub-adults, and four adults. We analyzed fecal samples using Terminal-Restriction Fragment Length Polymorphism (T-RFLP) of amplified 16S rRNA genes to determine bacterial diversity present. Between 1 and 14 terminal-restriction fragments (T-RFs) were observed in each sample. A total of 26 unique T-RFs were produced from the samples and ranged in size from 92 to 837 base pairs (bps). Twenty-four of these T-RFs corresponded to five bacterial phyla: Actinobacteria, Bacteroidetes, Firmicutes, Mollicutes, and Proteobacteria, as well as uncultured and unidentified bacterial species. The remaining T-RFs corresponded solely to uncultured or unidentified bacteria. Firmicutes was the most common phylum, observed in 11 of the samples. Bacteroidetes was the second-most common phylum, detected in 8 of the samples. Principal Components Analysis (PCA) revealed a discrete clustering of 10 samples when looking at components one and two, and a clustering of 11 samples when looking at component three. These three components accounted for 72.5% of the variation within the data. Morisita indices were computed to compare T-RF profiles of two samples at a time, and were between 0 and 0.886. Results indicated that some fecal bacterial profiles were similar in the study group, but ultimately varied between samples when compared two at a time. Specific diet, physiology, and environmental reservoir exposure may play large roles in shaping such profiles. Master of Science

This research relates Hg levels in the Common Loon (Gavia immer) to a variety of physical factors. Constructed within the framework of a GIS system, this model analyzes the spatial relationships and the influence of physical land cover factors as a function of distance from the individual loon territories. Thiessan polygons were used to generate the territory for each loon. Buffering of the thiessan polygons was done to establish the boundaries of the individual distance classes and to gather information on the percentage of individual land cover classes within each distance class. Information on precipitation was also gathered. Results from the regression analysis (R2 = 57.3% at the 150m distance class) performed on the variables suggest that the proximity of certain land use types such as cropland, shrub land, and wetlands influence the rates at which Hg is available within an individual territory. Within the 150m and 300m buffers, crop land, shrub land, and wetland exhibited the strongest relationship with the Hg levels in the common loon, with cropland exhibiting a negative relationship suggesting that the proximity of cultivated lands plays a role in decreasing the amount of available Hg in a territory. Master of Science