Cochlear pharmacokinetics - Micro-computed tomography and learning-prediction modeling for transport parameter...

Area of Science:

• Oto-pharmacology
• Biomedical Imaging
• Computational Biology

Background:

• Inner ear disorders, including sensorineural deafness and genetic diseases, present challenges for effective treatment.
• Current pharmacokinetic models rely on invasive methods, limiting spatial resolution and research scope.
• Developing noninvasive methods for inner ear drug delivery modeling is crucial.

Purpose of the Study:

• To develop a novel intracochlear pharmacokinetic model for the murine cochlea.
• To utilize an imaging, learning-prediction (LP) paradigm for determining drug transport parameters.
• To enable accurate, noninvasive assessment of drug concentrations within the inner ear.

Main Methods:

• Employed noninvasive micro-computed tomography (micro-CT) imaging during in vivo contrast agent infusion.
• Registered 3-D micro-CT scans to a cochlear atlas for precise segmentation and concentration extraction.
• Developed an LP model to learn concentration-dependent diffusion and inter-scala transport parameters.

Main Results:

• The LP model accurately captured spatio-temporal drug concentration profiles in the murine cochlea.
• Model performance was comparable to existing state-of-the-art pharmacokinetic models.
• Simulations demonstrated the model's ability to handle various infusion molecules and delivery protocols.

Conclusions:

• The developed intracochlear pharmacokinetic model offers a noninvasive approach for studying inner ear drug delivery.
• The model can simulate drug concentrations and optimize delivery protocols to minimize toxicity.
• The methodology is scalable to larger animals and diverse drug infusion strategies.