Reduction of Doubtful Detection of Human Lymphocyte Micro-Nucleus Based on Wiener’s Deconvolution and Computing Distributed Service

In order to perform further reduction on the number of rejected images by the cascade of the pattern matching analysis and the blob analysis for the correct detection of the Micro Nucleuses into human lymphocytes, the Wiener's deconvolution is proposed. The heavy computation burden compared with the previous procedure based on spatial filters suggests to use the Wiener 's deconvolution to process the rejected images, only, and not all the acquired images from the Image Flow Cytometer. According to this consideration and to speed up the complete pre-processing procedure, the implementation of the correction procedure is based on a computing distributed service implemented by using the LabVIEW Shared Variables (SV). Experimental tests confirm the fitness of the correction method and the advantages of the computing distributed service as regarding both the speed up and the uniformity of the analysis time.