Dynamic Analysis of Coupled Sawtooth and Rectangle Cellular Neural Networks with Application in HBV Patients' B-Scan Images

Liver damage caused by hepatitis B virus (HBV) infections is diffuse. Live medical images and FibroScan images show that it is inhomogeneous. Modeling and interpreting such phenomena are both important for theoretical research and practical application. The cellular neural networks (CNNs) introduced by Chua and Yang can model widely observed patterns in both nonbiological and biological media. The sawtooth and rectangle (SR) CNN introduced by Chua and colleagues is able to generate SR-shaped patterns for any random-input grayscale patterns, which are also similar to some chronic HBV-infected patients' liver B-scan images. This paper will mathematically study the dynamic behaviors of the SR CNN, numerically simulate the output images of the SR CNN, and biologically interpret the relationships between the output images of the SR CNN and the patterns in HBV-infected patients' liver B-scan images. Our research results show that the SR CNN may be used as a candidate for modeling liver infections.