Implementing A Dynamic Fuzzy State Estimator for Predicting M1-M2 Phenotypes Switch in Post Ischemic Stroke

The role of Microglia cells after ischemic stroke has been approved in the neuroinflammation procedure. However, the mechanism of switching activated microglia to M1-M2 phenotypes is still unknown to the scientists. This switching process contains an uncertain dynamic and non-linear system that most computational methods including probabilistic reasoning could be very difficult to handle modelestimation and further in-silico analysis. Materials and Methods: A dynamic fuzzy state estimator has been designed to estimate the M1-M2 states. Input data for training our model is the M1 microglia, which is activated by lipopolysaccharide produces proinflammatorycytokine and M2 phenotypes, which is activated by interleukin (IL-4) and expresses antiinflammatory mediators in the different stages. Fuzzyrules and inference engine are designed to setup our dynamic system within a k-fold training and testing sets. Results: In order to achieve a better understanding of the M1-M2 switching process after microglia activation, computational model within dynamic state estimators (finite states) is considered. In our implemented method, Fuzzy systems, is simulated with Gaussian membership functions and Mamdani Interpretation. The accuracy of the output model is also related to the input data which guarantees the computation robustness in estimation. Therefore, Predicting M1 and M2 microglia switch procedure could assist researchers to prevent Neurodegenerative Diseases. Conclusion: Our research demonstrates a computational model based on the fuzzy dynamic estimator. Early estimation mechanism may aid researchers to prevent neurodegenerative disease due to the neuroinflammation after ischemic stroke