In-Vivo Imaging of Cell Migration Using Contrast Enhanced MRI and SVM Based Post-Processing

Area of Science:

• Biomedical Imaging
• Nanotechnology
• Machine Learning

Background:

• Cell migration is crucial in biological processes but challenging to track non-invasively.
• Current MRI methods with iron oxide nanoparticles lack sensitivity and specificity for cell detection.

Purpose of the Study:

• To develop a quantitative, non-invasive in-vivo cell localization method.
• To improve sensitivity and specificity for detecting iron oxide nanoparticle-labeled cells.
• To enable prolonged monitoring of cell migration in living organisms.

Main Methods:

• Utilized contrast-enhanced multiparametric MRI combined with support vector machines (SVM) for post-processing.
• Trained and evaluated SVM using imaging phantoms with varying concentrations of iron oxide nanoparticle-labeled cancer cells.
• Extracted multiparametric features from T2*-weighted gradient-echo scans (magnitude, phase, relaxation rates, phase perturbations).

Main Results:

• Achieved high detection quality using SVM analysis of multiparametric feature-space.
• Demonstrated high specificity and sensitivity in phantoms, with a detection limit of 30 labeled cells/mm³ (19 μM iron oxide).
• Successfully applied the method to track labeled cancer cell migration in rats in vivo.

Conclusions:

• The developed method offers high spatial resolution, specificity, and sensitivity for non-invasive in-vivo cell detection.
• This approach is suitable for long-term cell migration studies.
• Combines iron oxide labeling, multiparametric MRI, and SVM for advanced cell tracking.