“
“Purpose: The objective

of this study was to prospectively verify if diffusion-weighted magnetic resonance (DwMR)-related parameters such as perfusion fraction (f) and slow diffusion coefficient (D), according to Le Bihan theory, are more effective than apparent diffusion coefficient (ADC) for classification and characterization of the more frequent focal liver lesions (FLLs) in noncirrhotic liver. Methods: Sixty-seven patients underwent standard liver magnetic resonance imaging (MRI) and free-breath multi-b DwMR study. Two regions of interest PKC412 purchase were defined by 2 observers, including 1 FLL for each patient (21 hemangiomas, 21 focal nodular hyperplasias, 25 metastases) and part of surrounding parenchyma, respectively. For every FLL, D, f, and ADC were estimated both as absolute value and as ratio between FLL and surrounding

parenchyma click here by fitting the reduced equation of the bicompartmental model to experimental data; t test, analysis of variance, and receiver operating characteristic analysis were performed. Results: t Test showed significant differences in ADC(lesion), f(lesion), D-lesion, ADC(ratio), and D-ratio values between benign and malignant FLLs, more pronounced for ADC(lesion) (P smaller than 0.0009) and ADC(ratio) (P = 0.001). Applying cutoff values of 1.55 x 10(-3) mm(2)/s (ADC(lesion)) and 0.89 (ADC(ratio)), the DwMR study presented sensitivities and specificities, respectively, of 84% and 80% (for SBI-0206965 mw ADC(lesion)), 72% and 80% (ADC(ratio)). Conclusions: Apparent diffusion coefficient (by fitting procedures) better performs than do D and f in FLL classification, especially when its values are less than 1.30 or greater than 2.00 x 10(-3) mm(2)/s.”
“This work aimed to compare the predictive

capacity of empirical models, based on the uniform design utilization combined to artificial neural networks with respect to classical factorial designs in bioprocess, using as example the rabies virus replication in BHK-21 cells. The viral infection process parameters under study were temperature (34 degrees C, 37 degrees C), multiplicity of infection (0.04, 0.07, 0.1), times of infection, and harvest (24, 48, 72 hours) and the monitored output parameter was viral production. A multilevel factorial experimental design was performed for the study of this system. Fractions of this experimental approach (18, 24, 30, 36 and 42 runs), defined according uniform designs, were used as alternative for modelling through artificial neural network and thereafter an output variable optimization was carried out by means of genetic algorithm methodology. Model prediction capacities for all uniform design approaches under study were better than that found for classical factorial design approach.