Motion Compensation in Pulmonary Fluorescence Lifetime Imaging: An Image Processing Pipeline for Artefact Reduction...

Area of Science:

• Biomedical optics
• Medical imaging analysis
• In-vivo imaging technologies

Background:

• Motion artefacts degrade image quality in real-time pulmonary Optical Endomicroscopy (OEM).
• Accurate image registration is crucial for reliable Fluorescence Lifetime Imaging (FLIm) data analysis in vivo.
• Existing methods struggle with real-time processing and motion correction for FLIm in OEM.

Purpose of the Study:

• Introduce Temporal Reliability and Accuracy via Correlation Enhanced Registration (TRACER), a novel pipeline for FLIm data in pulmonary OEM.
• Address motion artefacts to enhance accuracy and reliability of in-vivo FLIm image sequences.
• Develop a computationally efficient and robust image registration method.

Main Methods:

• TRACER pipeline integrates pre-processing (uninformative frame removal, dense optical flow for motion characterisation) and novel registration.
• Employs a tracking-based Normalised Cross Correlation registration using Channel and Spatial Reliability Tracker.
• Focuses on precise alignment of FLIm data acquired during in-vivo pulmonary OEM.

Main Results:

• TRACER pipeline demonstrated 20% to 30% performance enhancement across various metrics.
• The TRACER registration approach surpassed state-of-the-art methods in image registration.
• Achieved an order-of-magnitude faster runtime compared to the next best method.

Conclusions:

• TRACER provides a robust solution for motion artefact correction in FLIm data.
• The pipeline ensures improved image quality and real-time feasibility for in-vivo pulmonary OEM.
• TRACER enhances the reliability and accuracy of FLIm sequences for clinical applications.