Integrated state evaluation for the images of crystallization droplets utilizing linear and nonlinear classifiers

Abstract

In a usual crystallization process, the researchers evaluate the protein crystallization growth states based on visual impressions and repeatedly assign scores throughout the growth process. Although the development of crystallization robotic systems has generally realised the automation of the setup and storage of crystallization samples, evaluation of crystallization states has not yet been completely automated. The method presented here attempts to categorize individual crystallization droplet images into five classes using multiple classifiers. In particular, linear and nonlinear classifiers are utilized. The algorithm is comprised of pre-processing, feature extraction from images using texture analysis and a categorization process using linear discriminant analysis (LDA) and support vector machine (SVM). The performance of this method has been evaluated by comparing the results obtained using the method with the results obtained by a human expert and the concordance rate was 84.4%.

Simultaneous false‐colour imaging of birefringence, extinction and transmittance at camera speed

“…Milliview shows promise for large‐scale, automated protein crystal growth. Techniques developed previously, despite being able to detect proteins on the basis of their birefringence (Echalier et al , 2003; Nollert, 2003; Owen & Garman 2 , 2005; Kawabata et al , 2006), are either too slow for automatic screening of the large number of samples generated by the novel proteomic techniques or too inaccurate (Asundi et al , 2001).…”

Applications of Artificial Intelligence and Machine Learning Algorithms to Crystallization

“…Interestingly, several studies have reported the benefits of employing multiple classifiers simultaneously. Linear and nonlinear algorithms employed together (e.g., linear and radial basis-function kernel SVM together with ANNs or an LDA with an SVM model) have improved the classification of protein crystals. , …”

Automatic detection of crystallization evaluation region by using image processing

“…Three images for which the present method is inappropriate and resulting relationship of radius to variance distribution and r to mindif_r As our future work, we will design and develop the automatic crystallization growth evaluation software which consists of proposed method and our imaging-processing methods (Kawabata et al, 2006a(Kawabata et al, , b, 2008Saitoh et al, 2005).…”

“…Research on automated evaluation of the state of growth in crystallization solutions has involved the use of polarizing filters image processing, and various other techniques for the detection of the protein crystals in the crystallization well. We have proposed several methods for this purpose (Kawabata et al, 2006a(Kawabata et al, , b, 2008Saitoh et al, 2005). In any evaluation process, it is necessary to extract evaluation region from microscope images.…”