Multi-detector fusion and Bayesian smoothing for tracking viral and chromatin structures.

Area of Science:

• Biophysics
• Cell Biology
• Image Analysis

Background:

• Accurate tracking of intracellular and viral structures in fluorescence microscopy is crucial for quantifying cellular processes.
• Existing methods face challenges with varying spot sizes and complex cellular environments.

Purpose of the Study:

• To develop a novel probabilistic tracking approach for multiple particle tracking in fluorescence microscopy images.
• To enhance accuracy and robustness by integrating multi-detector, multi-scale data fusion, and Bayesian smoothing.

Main Methods:

• A probabilistic tracking approach utilizing multi-detector and multi-scale data fusion.
• An intensity-based covariance intersection method for fusing detections from various sources (classical and deep learning).
• Bayesian smoothing to integrate predictions from past and future time points.

Main Results:

• Achieved state-of-the-art results on the Particle Tracking Challenge dataset.
• Outperformed existing methods on challenging live-cell fluorescence microscopy data of viral and cellular proteins.
• Demonstrated superior performance on chromatin structure tracking in infected and non-infected cells.

Conclusions:

• The proposed probabilistic tracking approach offers superior performance compared to existing methods.
• The integration of data fusion and Bayesian smoothing provides a robust framework for particle tracking.
• This method advances the quantification of cellular processes through accurate microscopy image analysis.