A Cognitive Interpretation of Thermographic Images Using Novel Fuzzy Learning Semantic Memories

Abstract

AbstractFuzzy neural network (FNN), a hybrid of fuzzy logic and neural network, is computationally powerful, robust and able to model a complex nonlinear problem domain via the extraction and self-tuning of fuzzy IF-THEN rules. This yields a powerful semantic learning memory system that is useful to build intelligent decision support tools. Thermal imaging has been effectively used in the detection of infrared spectrum for the screening of potential SARS patients. This chapter proposes a cognitive approach in the study of the correlation and semantic interpretation of superficial thermal images against the true internal body temperature. Comparison between global and local semantic memories as well as Mamdani and TSK model of FNNs are presented. Existing infrared systems are commonly used at various boarder checkpoints and these have high false-negative rate. The use of FNN as a back-end of the system can significantly improves and hence serves the role of an intelligent medical decision support tool with high degree of accuracy. Extensive experimentations are conducted on real-life data taken from the Emergency Department (A&E), Tan Tock Seng Hospital (the designated SARS center in Singapore). The performance of FNN as the thermal analysis decision support system, providing plausible semantic interpretation and understanding, is highly encouraging.KeywordsSemantic interpretationlearning memoriesfuzzy rulesThermal imagingthermographsneural fuzzy networksintelligent decision support systemcognitive approachbrain-inspiredtemperature correlationlocal and global semantic networksMamdani and TSK models