Exploring fractal dimensions on ultrasound of gonadal images for sex determination in shortnose sturgeon (Acipenser brevirostrum)

Sturgeon are often viewed as monomorphic species because they often lack external features that allow differentiation between sexes. Current practices for sexing sturgeon rely heavily on surgical invasive procedures to visually examine gonadal tissues. An alternative which is widely used in different species of Acipenserids is the use of ultrasound to sex them. The challenge with ultrasound is that it requires an experienced operator to successfully sex the fish by recognizing relevant patterns in the structures of gonads. The objective of this study is to attempt to lay the groundwork for potential systematization of sexing techniques using a portable ultrasound as a medium. We used texture analysis software based on lacunarity measurements and fractal dimensions to determine whether male gonads, female granular tissue (GT) and female pinheads were significantly different from each other using two different resolutions on the ultrasound machine. Male gonads and GT were significantly different from pinheads in lacunarity measurements (p=0.001) using 10-12Mhz general resolution imaging. Fractal dimensions did not result in any significant differences. Lacunarity has the potential to determine sex based on gray level-co-occurrence matrices and the software is freely available. We have also developed a relative probability table based on the data gathered on lacunarity in this study available as supplementary material for quick reference.