The necessity for sophisticated and precise diagnostic instruments for the prompt recognition of COVID-19 patients has been highlighted by the continuing worldwide epidemic. In this regard, this study presents a unique method for accurately classifying X-ray images of chest in COVID-19 prediction by combining Neutrosophic Fuzzy Logic with a Hybrid CNN and LSTM architecture. Medical image analysis involves uncertainties and imprecise information, which is handled via Neutrosophic Fuzzy Logic. The suggested hybrid model offers a thorough comprehension of the spatial and temporal patterns in chest X-ray pictures by utilizing the advantages of CNN for feature extraction and LSTM for sequential information learning. Hybrid CNN-LSTM architecture based on Neutrosophic Fuzzy Logic is trained on an enormous set of various chest X-ray pictures, including both positive and negative instances of COVID-19 and other respiratory diseases. The proposed method is implemented using Python software. In addition to improving COVID-19 prediction accuracy, the combination of Neutrosophic Fuzzy Logic with a Hybrid CNN-LSTM structure creates a strong framework for managing uncertainty in medical image classification tasks. The proposed CNN-LSTM model with Neutrosophic Fuzzy logic shows better accuracy with 98.6% which is 4.4 % higher when compared with COVID CAPS , Bayesian CNN , Deep Feature + SVM and DCNN. This study represents a major advancement in the creation of sophisticated and trustworthy diagnostic instruments for effective healthcare administration during times of worldwide health emergencies.
Digital Object Identifier (DOI)
M. Mirza, Olfat and H. Samak, Ahmed
"Neutrosophic Fuzzy Logic-Based Hybrid CNN- LSTM for Accurate Chest X-ray Classification in COVID-19 PredictionNeutrosophic Fuzzy Logic-Based Hybrid CNN- LSTM for Accurate Chest X-ray Classification in COVID-19 Prediction,"
Applied Mathematics & Information Sciences: Vol. 18:
1, Article 14.
Available at: https://digitalcommons.aaru.edu.jo/amis/vol18/iss1/14