This paper presents a comprehensive review of retinal blood vessel segmentation based on deep learning. The geometric characteristics of retinal vessels reflect the health status of patients and help to diagnose some diseases such as diabetes and hypertension. The accurate diagnosis and timing treatment of these diseases can prevent global blindness of patients. Recently, deep learning algorithms have been rapidly applied to retinal vessel segmentation due to their higher efficiency and accuracy, when compared with manual segmentation and other computer-aided diagnosis techniques. In this work, we reviewed recent publications for retinal vessel segmentation based on deep learning. We surveyed these proposed methods especially the network architectures and figured out the trend of models. We summarized obstacles and key aspects for applying deep learning to retinal vessel segmentation and indicated future research directions. This article will help researchers to construct more advanced and robust models.

Alzheimer's disease (AD) is a neurodegenerative disorder that causes memory degradation and cognitive function impairment in elderly people. The irreversible and devastating cognitive decline brings large burdens on patients and society. So far, there is no effective treatment that can cure AD, but the process of early-stage AD can slow down. Early and accurate detection is critical for treatment. In recent years, deep-learning-based approaches have achieved great success in Alzheimer's disease diagnosis. The main objective of this paper is to review some popular conventional machine learning methods used for the classification and prediction of AD using Magnetic Resonance Imaging (MRI). The methods reviewed in this paper include support vector machine (SVM), random forest (RF), convolutional neural network (CNN), autoencoder, deep learning, and transformer. This paper also reviews pervasively used feature extractors and different types of input forms of convolutional neural network. At last, this review discusses challenges such as class imbalance and data leakage. It also discusses the trade-offs and suggestions about pre-processing techniques, deep learning, conventional machine learning methods, new techniques, and input type selection.

A self-body-biasing technique is proposed for differential-drive cross-coupled (DDCC) rectifier, with its profound application in far-field RF energy-harvesting systems. The conventional source-to-body, and the proposed technique known as Lower DC Feeding (LDCF), were fabricated in the 130-nm CMOS and compared at the operation frequency of 500 MHz, 953 MHz and 2 GHz along with a corresponding load of 2 kΩ, 10 kΩ and 50 kΩ. The technique allows the PMOS transistors to operate with a dynamic threshold voltage (Vth) which improves the power conversion efficiency (PCE) when the rectifier is operating at a smaller received power. A 9.5% of improvement is achieved at the peak PCE when the rectifier is operating at 953 MHz, and driving a 10 kΩ load. A maximum PCE of 73.9% is measured at 2 GHz when driving a 2-kΩ load. The LDCF technique also offers a self-limiting capability for its output voltage, by reducing the PCE at larger received power. A limit-voltage level of 3.5 V is measured irrespective to the operating frequency and load. This capability aids the protection of the subsequent circuits in a wireless sensor from being overpowered.

Cardiovascular disease (CVD) is the leading cause of death worldwide. A Machine Learning (ML) system can predict CVD in the early stages to mitigate mortality rates based on clinical data. Recently, many research works utilized different machine learning approaches to detect CVD or identify the patient’s severity level. Although these works obtained promising results, none focused on employing optimization methods to improve the ML model performance for CVD detection and severity-level classification. This study provides an effective method based on the Synthetic Minority Oversampling Technique (SMOTE) to handle imbalance distribution issue, six different ML classifiers to detect the patient status, and Hyperparameter Optimization (HPO) to find the best hyperparameter for ML classifier together with SMOTE. Two public datasets were used to build and test the model using all features. The results show that SMOTE and Extra Trees (ET) optimized using hyperband achieved higher results than other models and outperformed the state-of-the-art works by achieving 99.2% and 98.52% in CVD detection, respectively. Also, the developed model converged to 95.73% severity classification using the Cleveland dataset. The proposed model can help doctors determine a patient’s current heart disease status. As a result, it is possible to prevent heart disease-related mortality by implementing early therapy.