Diabetic Retinopathy (DR) is an eye condition that mainly affects individuals who have diabetes and is one of the important causes of blindness in adults. As the infection progresses, it may lead to permanent loss of vision. Diagnosing diabetic retinopathy manually with the help of an ophthalmologist has been a tedious and a very laborious procedure. This paper not only focuses on diabetic retinopathy detection but also on the analysis of different DR stages, which is performed with the help of Deep Learning (DL) and transfer learning algorithms. CNN, hybrid CNN with ResNet, hybrid CNN with DenseNet are used on a huge dataset with around 3662 train images to automatically detect which stage DR has progressed. Five DR stages, which are 0 (No DR), 1 (Mild DR), 2 (Moderate), 3 (Severe) and 4 (Proliferative DR) are processed in the proposed work. The patient’s eye images are fed as input to the model. The proposed deep learning architectures like CNN, hybrid CNN with ResNet, hybrid CNN with DenseNet 2.1 are used to extract the features of the eye for effective classification. The models achieved an accuracy of 96.22%, 93.18% and 75.61% respectively. The paper concludes with a comparative study of the CNN, hybrid CNN with ResNet, hybrid CNN with DenseNet architectures that highlights hybrid CNN with DenseNet as the perfect deep learning classification model for automated DR detection.

Wireless Sensor Networks (WSNs) are experiencing rapid growth, reshaping human lifestyles. They consist of numerous spatially distributed sensor nodes dedicated to monitoring and storing environmental data. WSNs have been a vital component of the Internet of Things (IoT) concept since its inception, where all sensor nodes typically connect to the internet for data exchange. However, WSNs lack a direct internet connection for their nodes, necessitating the use of mediators to establish connectivity. Substantial scholarly research in recent years has focused on integrating WSNs into the IoT. Wireless Sensor Networks consist of a large number of sensors but face constraints due to limited energy capacity. These nodes can perform effectively in challenging environmental conditions, but their batteries cannot be recharged or replaced in harsh environments and also will lead to problems in basis of cost. Consequently, energy preservation is crucial for the network’s sustainability. In the field of networking, routing protocols play a significant role in impacting energy consumption. The primary focus during the design of routing protocols is the optimization of energy utilization. This paper aims to minimize overall energy consumption and enhance the network’s lifespan using Cluster-based routing algorithms. Clusters, which are essentially groups of sensors, feature a central entity called the cluster head (CH), along with associated sensor nodes (SN). Cluster heads are high-energy nodes responsible for collecting and transmitting data from other nodes to the base station (BS). This research introduces a novel energy-efficient clustering approach based on the Nature Inspired Improved Squirrel Search Algorithm (ISSA) with specified objective functions. The selection of cluster heads considers factors like node residual energy, distance from neighbors and distance from the base station, node degree, and load based on packet processing. The performance of ISSA is compared with other optimization algorithms like SSA, GWO and LEACH. The effectiveness of this proposed approach had given better results in terms of metrics such as the number of active nodes, dead nodes, energy consumption, and data packets received by the BS.