Multidimensional classification and diagnosis of leukemia in flow cytometry

This research is to establish new optimization methods for pattern recognition and classification of different white blood cells in actual patient data to enhance the process of diagnosis. Beckman-Coulter Corporation supplied flow cytometry data of numerous patients that are used as training sets to exploit the different physiological characteristics of the different samples provided. The methods of Support Vector Machines (SVM) and Artificial Neural Networks (ANN) were used as promising pattern classification techniques to identify different white blood cell samples and provide information to medical doctors in the form of diagnostic references for the specific disease states, leukemia. The obtained results prove that when a neural network classifier is well configured and trained with cross-validation, it can perform better than support vector classifiers alone for this type of data. Furthermore, a new unsupervised learning algorithm---Density based Adaptive Window Clustering algorithm (DAWC) was designed to process large volumes of data for finding location of high data cluster in real-time. It reduces the computational load to ∼O(N) number of computations, and thus making the algorithm more attractive and faster than current hierarchical algorithms.

Distribution of Diatoms Along Environmental Gradients in the Charlotte Harbor, Florida (USA), Estuary and Its Watershed: Implications for Bioassessment of Salinity and Nutrient Concentrations

The relative abundance of diatom species in different habitats can be used as a tool to infer prior environmental conditions and evaluate management decisions that influence habitat quality. Diatom distribution patterns were examined to characterize relationships between assemblage composition and environmental gradients in a subtropical estuarine watershed. We identified environmental correlates of diatom distribution patterns across the Charlotte Harbor, Florida, watershed; evaluated differences among three major river drainages; and determined how accurately local environmental conditions can be predicted using inference models based on diatom assemblages. Sampling locations ranged from freshwater to marine (0.1–37.2 ppt salinity) and spanned broad nutrient concentration gradients. Salinity was the predominant driver of difference among diatom assemblages across the watershed, but other environmental variables had stronger correlations with assemblages within the subregions of the three rivers and harbor. Eighteen indicator taxa were significantly affiliated with subregions. Relationships between diatom taxon distributions and salinity, distance from the harbor, total phosphorus (TP), and total nitrogen (TN) were evaluated to determine the utility of diatom assemblages to predict environmental values using a weighted averaging-regression approach. Diatom-based inferences of these variables were strong (salinity R 2 = 0.96; distance R 2 = 0.93; TN R 2 = 0.83; TP R 2 = 0.83). Diatom assemblages provide reliable estimates of environmental parameters on different spatial scales across the watershed. Because many coastal diatom taxa are ubiquitous, the diatom training sets provided here should enable diatom-based environmental reconstructions in subtropical estuaries that are being rapidly altered by land and water use changes and sea level rise.