IDENTIFYING THE MORPHOLOGICAL GROWTH PATTERNS OF MICROBIOLOGICAL DATA TYPES USING COMPUTER-VISION AND STATISTICAL MODELLING
Sabeeha Sultana, Mohammad Basha · INTERNATIONAL JOURNAL OF COMPUTER APPLICATION · 2020
An automatic tool is industrialized to recognize and classify morphological growth patterns of various microbiological data types by means of computer-vision and statistical modelling techniques.In algae phage (phage) typing, representative profiles of morphological growth stages of different algae types are extracted.Present systems rely on the subjective reading of the profiles of the by a human expert, which is time-consuming and prone to errors.The statistical methodology presented in this work, provides for an automated, objective and robust analysis of the visual data, along with the ability to cope with increasing data volumes.Validation is performed by comparison to an expert manual segmentation and labelling of the phage profiles.The statistical analysis such as the descriptive statistics is performed on time series data extracted is important for understanding relationships between parameters, provides insight into the growth curve of microalgae and cyanobacteria (correlation) and an essential step to estimate yield of biomass, etc., or predict the duration to achieve a certain yield of a pigment or protein, etc., for commercial applications.There are numerous methods for modelling time series data and being able to predict specific values and ; specifically, Regression Analysis and Analysis of Variance (ANOVA) are foremost among them.Computation of the correlation coefficient aids in better understanding the relationships that exist between various parameters that evolve with time and change with different phases of the growth of the organism (and cyanobacteria).This study focuses on statistical techniques for the analysis of time-series data.