Performance evaluation of spot detection algorithms in fluorescence microscopy images

Abstract

Detection of messenger Ribonucleic Acid (mRNA) spots in fluorescence microscopy images is of great importance for biologists seeking better understanding of cell functionality. Fluorescence microscopy and specific staining methods make biological molecules appear as bright spots in image data. Manual analysis of such data is both time-consuming and laborious, and can lead to errors. In this study, we compare two computer-based detection methods used in fluorescence microscopy images for the detection of spots. The algorithms compared are the Isotropic Undecimated Wavelet Transform (IUWT) and the Feature Particle Detection (FPD). The performance of these algorithms is validated on synthetic images. Our study finds a major difference in the performance of the two algorithms. IUWT performs better than FPD.