No abstract

Skin cancer is one the most dangerous types of cancer and is one of the primary causes of death worldwide. The number of deaths can be reduced if skin cancer is diagnosed early. Skin cancer is mostly diagnosed using visual inspection, which is less accurate. Deep-learning-based methods have been proposed to assist dermatologists in the early and accurate diagnosis of skin cancers. This survey reviewed the most recent research articles on skin cancer classification using deep learning methods. We also provided an overview of the most common deep-learning models and datasets used for skin cancer classification.

The Internet of Things (IoT) and the integration of medical devices perform hand-to-hand solutions and comfort to their users. With the inclusion of IoT under medical devices a hybrid (IoMT) is formulated. This features integrated computation and processing of data via dedicated servers. The IoMT is supported with an edge server to assure the mobility of data and information. The backdrop of IoT is a networking framework and hence, the security of such devices under IoT and IoMT is at risk. In this article, a framework and prototype for secure healthcare application processing via blockchain are proposed. The proposed technique uses an optimized Crow search algorithm for intrusion detection and tampering of data extraction in IoT environment. The technique is processed under deep convolution neural networks for comparative analysis and coordination of data security elements. The technique has successfully extracted the instruction detection from un-peer source with a source validation of 100 IoT nodes under initial intervals of 25 nodes based on block access time, block creation, and IPFS storage layer extraction. The proposed technique has a recorded performance efficiency of 92.3%, comparable to trivial intrusion detection techniques under Deep Neural Networks (DNN) supported algorithms.

In recent times, the usage of modern neuromorphic hardware for brain-inspired SNNs has grown exponentially. In the context of sparse input data, they are undertaking low power consumption for event-based neuromorphic hardware, specifically in the deeper layers. However, using deep ANNs for training spiking models is still considered as a tedious task. Until recently, various ANN to SNN conversion methods in the literature have been proposed to train deep SNN models. Nevertheless, these methods require hundreds to thousands of time-steps for training and still cannot attain good SNN performance. This work proposes a customized model (VGG, ResNet) architecture to train deep convolutional spiking neural networks. In this current study, the training is carried out using deep convolutional spiking neural networks with surrogate gradient descent backpropagation in a customized layer architecture similar to deep artificial neural networks. Moreover, this work also proposes fewer time-steps for training SNNs with surrogate gradient descent. During the training with surrogate gradient descent backpropagation, overfitting problems have been encountered. To overcome these problems, this work refines the SNN based dropout technique with surrogate gradient descent. The proposed customized SNN models achieve good classification results on both private and public datasets. In this work, several experiments have been carried out on an embedded platform (NVIDIA JETSON TX2 board), where the deployment of customized SNN models has been extensively conducted. Performance validations have been carried out in terms of processing time and inference accuracy between PC and embedded platforms, showing that the proposed customized models and training techniques are feasible for achieving a better performance on various datasets such as CIFAR-10, MNIST, SVHN, and private KITTI and Korean License plate dataset.

No abstract