Medical Image Retrieval using Optimization Algorithm and Integration of Deep Learning Network (DLN) with Fuzzy Logic (FL)

Medical image safety and confidentiality are essential because of their highly sensitive nature and the severe implications that could result through unauthorized changes, such as data breaches and incorrect diagnoses. In order to match and score a query image based on similarity, a retrieval engine uses feature vectors, which are high-level image representations maintained by a Secure Medical Image Retrieval (SMIR) system. SMIR is used for medical image retrieval, processing semantic information or the same item for different class labels using one and numerous input queries. Because image search is ambiguous, it might be difficult to evaluate the image being searched across multiple image sources in order to optimize its retrieval. In order to facilitate the early identification and categorization of lung disorders, this research work proposed hybrid optimization approach. The intangible development of the conventional Artificial Gorilla Troops Optimization (AGTO) and Seagull Optimization Algorithm (SOA) is proposed. Finally, the suggested SMIR system uses a Hybrid Deep Learning Network (DLN) with Fuzzy Logic (FL) approach to identify and classify variables specific to lung diseases, using the ability to predict of deep neural models. Utilizing the elcap test image database, the suggested technique's effectiveness is evaluated. The accuracy score for suggested SMIR-hybrid AGTO with SOA - hybrid DLN with FL approach is 0.90 to classify the lung diseases for the given dataset with most accurate while compare with pro-classifier, CNN, RNN and BiLSTM model.