Adaptive Spot Detection With Optimal Scale Selection in Fluorescence Microscopy Images.

Area of Science:

• * Cell Biology
• * Biophysics
• * Image Analysis

Background:

• * Accurate detection of subcellular particles in fluorescence microscopy is crucial for quantitative analysis like counting and tracking.
• * Existing methods often require manual parameter tuning and struggle with variations in particle size and image background.
• * Vesicle segmentation is a key challenge in understanding cellular processes.

Purpose of the Study:

• * To develop an automated method for segmenting vesicles of similar sizes in fluorescence microscopy images.
• * To introduce a novel adaptive thresholding technique that eliminates the need for manual parameter tuning.
• * To provide a robust and computationally efficient solution for vesicle detection.

Main Methods:

• * Adaptive Thresholding of Laplacian of Gaussian (LoG) images with Auto-selected Scale (ATLAS) method.
• * Automatic selection of optimal scale based on vesicle size distribution using four criteria in a scale-space framework.
• * Locally adapted thresholding using a user-specified probability of false alarm (PFA) and local image statistics.

Main Results:

• * ATLAS automatically determines the optimal scale for vesicle detection, corresponding to the most frequent spot size.
• * The method achieves superior segmentation performance compared to existing techniques across multiple benchmark datasets.
• * ATLAS demonstrates robustness on real total internal reflection fluorescence microscopy images and requires minimal computation time.

Conclusions:

• * ATLAS provides an accurate, automated, and efficient solution for vesicle segmentation in fluorescence microscopy.
• * The method's adaptive thresholding and auto-selected scale eliminate the need for manual parameter tuning, enhancing usability.
• * ATLAS offers a significant advancement for quantitative analysis in cell biology and related fields.