An image analysis based expert system for assessing the quality of freeze-dried protein formulations

Introduction: Development of freeze-dried protein formulations is a complex and time consuming task, complicated by the numerous excipients available, the choice of processing conditions and the number of quality attributes to evaluate. The aim of this study was to develop a high throughput screening system for evaluating the visual appearance of the freeze-dried cake based on image analysis and innovative knowledge management solutions. Materials and methods: Samples consisted of different concentrations of mannitol, sucrose, model protein and dyes. 100 µL of sample solution was filled into each well of a brass well-plate sealed on one side. After freeze drying, images were obtained and analyzed with regard to color, grayscale value, sample size and a measure of the amount of information in the image termed image-entropy. Results and discussion: Different visual characteristics were evaluated using image analysis. Cake collapse was assessed by studying sample size (shrinkage), glassiness (transparency) was evaluated through the grayscale value, surface uniformity was determined through the image-entropy value of the well, and off-coloring was analyzed with color analysis. Although these sample attributes are all important parameters, none of them could be used alone, as visual appearance as observed by a specialist consists of an intricate combination of all this information. To cope with the complexity of visual assessment, an expert system was constructed to classify the samples according to overall visual appearance. Conclusion: Quality assessment of freeze-dried cakes was effectively performed using a high throughput image analysis based expert system.