Texture Analysis and Ultra high-frequency ultrasound in use for diagnosing Hirschsprung's disease

Abstract: An investigation into the use of statistical texture analysis in combination with ultra high-frequency ultrasound for diagnosing Hirschsprung’s disease (HD); An illness that manifests itself through the absence of ganglia cells in lower parts of the intestine. Current diagnostic techniques carry a higher risk of long-term consequences in newborns and children which enables a new avenue of research into new methods. This report delves into the use cases of the three statistical distribution types nakagami, skewness, and kurtosis. It has been found that these statistical texture parameters can differentiate between aganglionic and ganglionic tissues with high accuracy in longitudinal imaging and partly in transversal imaging. An interesting finding was that there is a difference between transversal and longitudinal images using Nakagami-m, which most likely is attributed to the direction of the muscle fibers. In longitudinal imaging, there is a significance of more than 99.9\% when using the ratio and the difference between aganglionic and ganglionic tissues. Additionally, it is shown that our method can be used both for diagnostics and during surgery. However, further research is needed to validate these findings along a wide range of transducers and develop a reliable diagnostic method for HD using ultrasound