Characterization and discrimination of spike protein in SARS-CoV-2 virus-like particles via surface-enhanced Raman spectroscopy.

Non-infectious virus-like particles (VLPs) are excellent structures for development of many biomedical applications such as drug delivery systems, vaccine production platforms, and detection techniques for infectious diseases including SARS-CoV-2 VLPs. The characterization of biochemical and biophysical properties of purified VLPs is crucial for development of detection methods and therapeutics. The presence of spike (S) protein in their structure is especially important since S protein induces immunological response. In this study, development of a rapid, low-cost, and easy-to-use technique for both characterization and detection of S protein in the two VLPs, which are SARS-CoV-2 VLPs and HIV-based VLPs was achieved using surface-enhanced Raman spectroscopy (SERS). To analyze and classify datasets of SERS spectra obtained from the VLP groups, machine learning classification techniques including support vector machine (SVM), k-nearest neighbors (kNN), and random forest (RF) were utilized. Among them, the SVM classification algorithm demonstrated the best classification performance for SARS-CoV-2 VLPs and HIV-based VLPs groups with 87.5% and 92.5% accuracy, respectively. This study could be valuable for the rapid characterization of VLPs for the development of novel therapeutics or detection of structural proteins of viruses leading to a variety of infectious diseases.

Highly Accurate Identification of Bacteria's Antibiotic Resistance Based on Raman Spectroscopy and U-Net Deep Learning Algorithms.

Bacterial pathogens especially antibiotic-resistant ones are a public health concern worldwide. To oppose the morbidity and mortality associated with them, it is critical to select an appropriate antibiotic by performing a rapid bacterial diagnosis. Using a combination of Raman spectroscopy and deep learning algorithms to identify bacteria is a rapid and reliable method. Nevertheless, due to the loss of information during training a model, some deep learning algorithms suffer from low accuracy. Herein, we modify the U-Net architecture to fit our purpose of classifying the one-dimensional Raman spectra. The proposed U-Net model provides highly accurate identification of the 30 isolates of bacteria and yeast, empiric treatment groups, and antimicrobial resistance, thanks to its capability to concatenate and copy important features from the encoder layers to the decoder layers, thereby decreasing the data loss. The accuracies of the model for the 30-isolate level, empiric treatment level, and antimicrobial resistance level tasks are 86.3, 97.84, and 95%, respectively. The proposed deep learning model has a high potential for not only bacterial identification but also for other diagnostic purposes in the biomedical field.

Design of a Remote Real-Time Monitoring System for Multiple Physiological Parameters Based on Smartphone.

Background: Utilization of the widely used wearable sensor and smartphone technology for remote monitoring represents a healthcare breakthrough. This study aims to design a remote real-time monitoring system for multiple physiological parameters (electrocardiogram, heart rate, respiratory rate, blood oxygen saturation, and temperature) based on smartphones, considering high performance, autoalarm generation, warning transmission, and security through more than one method. Methods: Data on monitoring parameters were acquired by the integrated circuits of wearable sensors and collected by an Arduino Mega 250 R3. The collected data were transmitted via a Wi-Fi interface to a smartphone. A patient application was developed to analyze, process, and display the data in numerical and graphical forms. The abnormality threshold values of parameters were identified and analyzed to generate an autoalarm in the system and transmitted with data to a doctor application via a third-generation (3G) mobile network and Wi-Fi. The performance of the proposed system was verified and evaluated. The proposed system was designed to meet main (sensing, processing, displaying, real-time transmission, autoalarm generation, and threshold value identification) and auxiliary requirements (compatibility, comfort, low power consumption and cost, small size, and suitability for ambulatory applications). Results: System performance is reliable, with a sufficient average accuracy measurement (99.26%). The system demonstrates an average time delay of 14 s in transmitting data to a doctor application via Wi-Fi compared with an average time of 68 s via a 3G mobile network. The proposed system achieves low power consumption against time (4 h 21 m 30 s) and the main and auxiliary requirements for remotely monitoring multiple parameters simultaneously with secure data. Conclusions: The proposed system can offer economic benefits for remotely monitoring patients living alone or in rural areas, thereby improving medical services, if manufactured in large quantities.