Evaluation of methods for detection of fluorescence labeled subcellular objects in microscope images

Abstract

BackgroundSeveral algorithms have been proposed for detecting fluorescently labeled subcellular objects in microscope images. Many of these algorithms have been designed for specific tasks and validated with limited image data. But despite the potential of using extensive comparisons between algorithms to provide useful information to guide method selection and thus more accurate results, relatively few studies have been performed.ResultsTo better understand algorithm performance under different conditions, we have carried out a comparative study including eleven spot detection or segmentation algorithms from various application fields. We used microscope images from well plate experiments with a human osteosarcoma cell line and frames from image stacks of yeast cells in different focal planes. These experimentally derived images permit a comparison of method performance in realistic situations where the number of objects varies within image set. We also used simulated microscope images in order to compare the methods and validate them against a ground truth reference result. Our study finds major differences in the performance of different algorithms, in terms of both object counts and segmentation accuracies.ConclusionsThese results suggest that the selection of detection algorithms for image based screens should be done carefully and take into account different conditions, such as the possibility of acquiring empty images or images with very few spots. Our inclusion of methods that have not been used before in this context broadens the set of available detection methods and compares them against the current state-of-the-art methods for subcellular particle detection.

Computer vision for image-based transcriptomics

“…5C). This seemingly minor, but essential, detail of our imageanalysis pipeline contrasts the most common high-throughput implementation of spot detection algorithms, which rescales the intensities of any image according to the intensities of its dimmest and brightest pixel [17,28,33]. While the accompanying code supports additional refinement of the spot detection, these additional parameters (2D/3D, minimal intensity of pixels, size of spots) have a negligible effect on the detection of transcripts once robust imaging conditions have been established experimentally.…”

Performance and sensitivity evaluation of 3D spot detection methods in confocal microscopy

“…Some studies, such as (21,35), used approximations, placing Gaussian intensity profiles directly in the image to simulate the spots. Poisson noise was then added as the dominant type of noise for fluorescent images.…”

“…Experiments were performed on synthetic image data as well as on real fluorescence microscopy images. Ruusuvuori et al (35) studied eleven spot detection methods. This study used widefield microscope images from well plate experiments with a human osteosarcoma cell line, as well as 2D slices from image stacks of yeast cells.…”

Adaptive Local Thresholding for Co-Localization Detection in Multi-Channel Fluorescence Microscopic Images

“…Ruusuvuori et al [7] adopted the band-pass filtering (BPF) to enhance the spots and suppress shot noises generated at the image acquisition phase. The feature point detection (FPD) method proposed by Sbalzarini and Koumouysakos [8] removes the background noise with a boxcar average and enhances the spots by a convolution with a Gaussian kernel.…”

“…This method was applied to detect membrane receptors in dendritic spines of mouse cerebellar Purkinje cells. The detection performance of this method was assessed by comparing with the five above mentioned, pre-existing methods for spot detection: BPF [7], FPD [8], HD [11], MW [9], SPL [10].…”