Multiway Principal Component Analysis for Smartphone Captured Image of Colloidal Gold Strip

Abstract:

Lateral flow test kit has become a popular device for immunoassay of biologically important molecules. Colloidal gold strip has been commercially available for point-of-care clinical testing at hospital or simply at home for self-testing. A method for reading the reaction results of colloidal gold strip assays is proposed. The RGB (red-green-blue) image of a reacted strip is captured by smartphone under natural-light illumination. The two-dimensional grayscale image of each color channel is transformed into nine images by spatially shifting the image in eight adjacent directions. After all three color channel images are transformed, an array including twenty-seven images is constructed. Multiway principle component analysis is then used to the array, extracting texture information of the original image. When the extracted texture is added back to the original image, color of control line and test line regions can be effectively enhanced for reading by naked eyes.

• Periodical: Advanced Materials Research (Volume 1035)
• Pages: 483-487
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.1035.483
• Online since: October 2014
• Authors: Lei Zhou, Yan Zhen Liu*, Chao Zhang
• Keywords: Colloidal Gold Strip, Image Processing, Lateral Flow Assay Component, Multiway PCA, Point-of-Care-Test

References

[1] Saied Assadollahi, Christiane Reininger, et al, Sensors 9(2009) 6084-6100.

[2] Suhyeon Kim, Je-Kyun Park, Biotechnology and Bioprocess Engineering 9( 2004) 127-131.

[3] Bharati H Manish, MacGregor F John, Texture analysis of images using Principal Component Analysis. Proceedings of SPIE. 4188(2001) 27-37.

[4] J M Prats-Montalban, A de Juan, et al, Chemometrics and intelligent laboratory systems, 107(2011) 1-23.

[5] P GELADI, H GRAHN, Multivariate Image Analysis, Encyclopedia of Analytical Chemistry, (2000) 85-117.