Image Processing of Finite Size Rat Retinal Ganglion Cells Using Multifractal and Local Connected Fractal Analysis

Automated image processing aids in classification of biological images. Many natural structures such as neurons may be multifractal and therefore not analyzable using current methods. The multifractal spectrum proposed here may mitigate this, Here we report the outcome of applying three methods that elucidate the variation within 16 rat retinal ganglion cells using the local connected fractal dimension (LCFD), mass-radius (MR) and maximum likelihood multifractal (MLM) analyses. Our results based on LCFD indicate that the neurons studied are possibly multifractal. However utilizing the MR method provided inconclusive results due to the finite size of the cells and the density variation throughout their structure. This has been addressed by utilizing a novel unbiased method-the MLM method. To improve the our results we are now aiming to use AI algorithms to optimize the selection of parameter values associated with the MLM method. 1. Introduction Fractal analysis is a useful tool in automated image processing, as it provides objective, quantitative measures that can help to characterize complex shapes. The aim of this work is to use such measures to increase our knowledge about the structure of neurons. Neurons are known to fall into several types, but distinguishing these types is a continuing problem and can be approached from an AI perspective. As neurons possess fractal structure, the global fractal dimension has been suggested as a useful measure. 3 Calculating the global fractal dimension of rat retinal ganglion cells has provided valuable additional data for classification of these cells and elucidating functional relationships. 4 Our work on rat RGCs suggests that the structure of such tissue is complex, and that there is great benefit to be obtained by applying a more sophisticated analysis than the global fractal dimension such as