Variability in a Starch Isolation Method and Automated Digital Image Analysis System Used for the Study of Starch Size Distributions in Wheat Flour

A starch isolation method and digital image analysis system were developed to accurately measure size distributions of the starch populations in wheat. The image analysis system was coupled directly to a light microscope equipped with computer controlled step stage and automatic focus. Automation of data acquisition and processing eliminated some of the labor intensive steps previously required for analyzing starch granule size distributions. This system was used to standardize starch isolation methods and compare variation and precision of the system. Operational variations were determined and statistically assessed. The number of fields of view required for low standard errors and acceptable speed of analysis was determined to be fifty. A major advantage of the system has been the increased resolution. The use of higher magnifications and stage automation allowed the analysis of starch granules as small as 0.84 μm in diameter while analyzing thousands of starch granules per sample.