Abstract:
An important worldwide health issue is pneumonia, a potentially fatal respiratory
illness. For a patient to receive appropriate treatment and care, pneumonia must be
promptly and accurately detected. Recent years have seen the emergence of promising
tools for automated pneumonia detection from chest radiographs, including medical
image analysis techniques and machine learning algorithms. The deep convolutional
neural network (Deep-CNN) multimodal model and transfer learning techniques are
utilised to construct and evaluate a machine learning application for pneumonia
detection in this abstract.The suggested methodology makes use of Deep-CNNs&
Transfer Learning Model combination, which have proven to perform exceptionally
well in image analysis applications. By using a multimodal approach, the model
makes use of both the contextual information and visual data retrieved from chest
radiographs, improving its capacity to identify significant patterns and features
suggestive of pneumonia. Additionally, transfer learning strategies are used to exploit
pre-trained models, giving the network access to information gained from sizable
datasets even in the absence of a substantial amount of labelled data.Experimental
results show that VIYU the state-of-the-artrnodel attained the highest accuracy and
recall score of98.08% and 98.91 %, respectively.
