Chaotic Sea Horse Optimization with Deep Learning Model for lung disease pneumonia detection and classification on chest X-ray images

Pneumonia is an acute respiratory illness caused by viruses or bacteria. Early detection of pneumonia is important to ensure curative treatment and improve survival rates. Pneumonia detection on chest X-rays (CXR) is important for early diagnosis, effective treatment, monitoring patient progress, and managing public health concerns. It plays a vital role in ensuring that individuals with pneumonia receive the appropriate care they need while contributing to research and disease surveillance efforts. However, the examination of CXRs is a difficult process and is prone to subjective variabilities. The use of artificial intelligence (AI) and deep learning (DL) models can perform the detection and classification of pneumonia on CXR images. With this motivation, this study introduces a new Chaotic Sea Horse Optimization with Deep Learning Method for Pneumonia Detection and Classification (CSHODL-PDC) technique on CXR images. The main intention of the CSHODL-PDC algorithm lies in the automated detection and classification of pneumonia on CXR images. The CSHODL-PDC method initially designs a Gaussian filtering (GF) based noise eradication approach to eliminate the noise. In addition, the CSHODL-PDC technique employs the NASNetLarge model to produce a set of feature vectors. Moreover, an improved fuzzy deep neural network (FDNN) model is applied for the automated identification and classification of pneumonia. Finally, the CSHO algorithm selects the optimal hyperparameter values of the improved FDNN model, demonstrating the novelty of the work. A series of simulation analyses were performed on the CXR Pneumonia dataset from the Kaggle repository. The experimental values inferred the improved performance of the CSHODL-PDC method over recent models with a maximum accuracy of 99.22%, precision of 98.96%, and recall of 99.22%. Therefore, the proposed model can be employed for accurate and automated pneumonia detection.