Quantifying barley morphology using the Euler characteristic transform

Anonymous

Keywords: Topological Data Analysis, Euler Characteristic Transform, plant morphology, barley inflorescence

TL;DR: Topological shape descriptors reveal "hidden" morphological information in barley seeds

Abstract: Shape plays a fundamental role in biology. Throughout history, the constant observation and documentation of the shape of multiple organisms has been key to further biological understanding of organism and tissue behavior, function, and origins. To extract, compare, and analyze such data diversity we must wield a quantifiable, robust, and concise method. We thus turn to Topological Data Analysis (TDA) and the Euler Characteristic Transform (ECT). As a study case, we quantify the morphology of X-ray CT scans of barley spikes and seeds using both traditional and topological shape descriptors based. We then successfully train a support vector machine to distinguish and classify 28 different parental genotypes of barley based solely on the 3D shape of their grains. We observe that combining both traditional and topological descriptors produces considerably better classification results compared to the use of exclusively traditional descriptors. This improvement suggests that TDA is thus a powerful complement to describe comprehensively a multitude of shape nuances which are otherwise not picked up by traditional morphometrics methods.