Surface measurement of the human lymphocyte micro-nucleus in image affected by alterations

In order to relieve structural chromosome aberrations, a new architecture of flow cytometer measurement device able to recognize and automatic count the Micro Nucleuses (MNs) on the acquired images of human lymphocytes has been pointed out. The technique adopted to perform the MN detection is based on the pattern matching algorithm. Biologist experts recommend that the correct detection of MN must respect the condition that the MN's surface is in the range of [1/3, 1/16] of the nucleus. The pattern matching algorithm can give only a coarse evaluation of the MN dimension. On the basis of these considerations, the paper proposes to cascade the pattern matching algorithm and the blob analysis in order to achieve the accurate evaluation of the MN surface. In order to evaluate the valid gray level used to threshold the image the pattern matching algorithm furnishes: (i) the coordinate (x, y) and (ii) the parameter about the coarse evaluation of MN dimension. So the blob analysis can be applied and the surfaces of both the MN and the nucleuses are accurately evaluated. ©2007 IEEE.