 Hi, I'm David Barry. Welcome to this video abstract in which I'm going to give an overview of the work that I've been doing using fractal geometry to analyze my serial structures. The motivation behind this work was to add to the existing audio knowledge showing the links between productivity and morphology and fermentation processes. And of course many reproducible relationships have been derived showing, for example, strong correlations between pellet size and metabolite yield in many processes. The problem or the limitation with these relationships is that they often overlook what's happening at the microscopic level and considering the evidence for a metabolite secretion occurring primarily at high-fold tips, this in my opinion represents a significant emission. Of course there have been a number of studies which have considered what's happening at the at the microscopic level for example, of various image analysis tools have been developed to automatically quantify the high-fold growth unit. The problem here is that in order to quantify the high-fold growth units freely dispersed high-fay have to be present such that the individual high-fay can be isolated and counted. Where larger aggregates are present the morphological parameters used are typically a projected area of circularity and various other interpretations thereof and of course these reveal very little about the underlying branching behavior. It also means that different metrics are used to quantify different morphologies and as such direct comparisons across different morphological classes are typically not possible. Well there have been a number of studies which have shown that filamentous microbes can produce what are approximately fractal structures and the big advantage of fractal analysis from a morphological quantification point of view is that it can be applied regardless of the morphological form that results in a particular process. There have also been a limited number of studies which have shown that branching behavior can influence fractal parameters. So I was interested in exploring this relationship further to see whether fractal analysis could be used to infer branching behavior within mycelial structures in cases where the hyphal growth unit could not be determined directly. Well in order to derive relationships between branching behavior and fractal parameters we obviously have to know what the branching behavior is and as just discussed this can be quite difficult to determine experimentally. So we can work around this by simulating the growth of mycelia. So this results in us having a large number of mycelial structures with known growth parameters and we can then release branching frequency for example with measured morphological parameters. So I used a simple model to simulate the growth of a large number of mycelia and at each stage during their development they were quantified according to two different fractal parameters. First being Lassenarity and the second being something of a term the Fourier fractal dimension which varies according to the shake of the mycelial boundary. So this constituted a training set within which we had known branching behaviors and fractal parameters. Second population of mycelia were then grown with random branching behaviors. They were quantified in the same way at each stage of their development. The fractal parameters that resulted were presented to the training set and an estimate of branching behavior returned. What I found was when averaging over a population of mycelia the estimated branching behaviors correlate them very strongly with the actual branching behaviors. So this fractal analysis represents potentially a very useful tool in morphological quantification and it can be applied regardless of the morphological form that results in a given process. That's everything. Thanks for watching. I hope you find the paper interesting. I hope you find it useful. I will be making the source code I developed available online relatively soon and of course if you have any questions on the paper please do contact me.