ALFRED: a novel image analysis application to inform mathematical modelling of microtubule networks in nerve cells
by
Sala de Conferências
Departamento de Física
Axons are the long, cable-like processes of neurons which form the nerves that wire our bodies. These delicate structures usually need to survive for an organism's lifetime (up to a century in humans) whilst being key lesion sites in injury (e.g. spinal cord injury), neurodegeneration (e.g. motorneuron disease) or age-related nervous system decay. We are particularly interested in analyses of continuous parallel bundles of filamentous polymers called microtubules (MTs), which run all along the axonal core providing the structural backbones. In ageing and neurodegeneration, axons can form swellings with disorganised MTs that have lost their bundled conformation, causing detrimental accumulations of organelles and vesicles.
Understanding the causes for such MT disorganisation requires their qualitative and quantitative description over time, i.e. the definition of informative parameters and development of suitable strategies to measure them reliably and efficiently.
The combined algorithms for measurement of parameters such as length, straightness or curvature were implemented computationally in a user-friendly MATLAB application ALFRED (Advanced Labelling, Fitting, Recognition and Enhancement of Data), which is able to import many image formats, skeletonise these images and apply our new analysis strategies. Our evaluations using well-defined patterns show that ALFRED measures their parameters within statistical significance, with impressive speed, and independent of image resolution.
Filipe Veloso e Pedro Costa