Smart Low Level Laser Therapy System for Automatic Facial Dermatological Disorder Diagnosis.

Computer-aided diagnosis using dermoscopy images is a promising technique for improving the efficiency of facial skin disorder diagnosis and treatment. Hence, in this study, we propose a low-level laser therapy (LLLT) system with a deep neural network and medical internet of things (MIoT) assistance. The main contributions of this study are to (1) provide a comprehensive hardware and software design for an automatic phototherapy system, (2) propose a modified-U2Net deep learning model for facial dermatological disorder segmentation, and (3) develop a synthetic data generation process for the proposed models to address the issue of the limited and imbalanced dataset. Finally, a MIoT-assisted LLLT platform for remote healthcare monitoring and management is proposed. The trained U2-Net model achieved a better performance on untrained dataset than other recent models, with an average Accuracy of 97.5%, Jaccard index of 74.7%, and Dice coefficient of 80.6%. The experimental results demonstrated that our proposed LLLT system can accurately segment facial skin diseases and automatically apply for phototherapy. The integration of artificial intelligence and MIoT-based healthcare platforms is a significant step toward the development of medical assistant tools in the near future.

Raspberry Pi Platform Wireless Sensor Node for Low-Frequency Impedance Responses of PZT Interface.

A wireless impedance monitoring system, called SSeL-Pi, is designed to have cheap, mobile, and handy practical features as compared to wired commercial impedance analyzers. A Raspberry Pi platform impedance sensor node is designed to measure signals at a low-frequency range of up to 100 kHz. The low-frequency impedance measurement via the proposed node has been combined with a new PZT interface technique for measuring local responses sensitive to structural damage. The new PZT interface can work as a surface-mounted or embedded sensor, and its local dynamic characteristics are numerically analyzed to pre-determine an effective impedance resonant frequency range of less than 100 kHz. Next, a software scheme was designed to visualize the input/output parameters of the proposed SSeL-Pi system (i.e., Raspberry Pi platform and PZT interface) and automate signal acquisition procedures of the impedance sensor node. The calibration for impedance signals obtained from the proposed system was performed by a series of procedures, from acquiring real and imaginary impedance to adjusting them with respect to a commercial impedance analyzer (HIOKI-3532). The feasibility of the wireless impedance monitoring system was experimentally evaluated for PZT interfaces that were subjected to various compressive loadings. The consistent results analyzed from signals measured by the SSeL-Pi and HIOKI 3532 systems were observed. Additionally, the strong relationships between impedance features (frequency shift and RMSD index) and compressive stresses of the PZT interfaces showed the potential for axial force/stress variation monitoring in real structures using the Raspberry Pi platform impedance sensor node and developed PZT interface.

Corroded Bolt Identification Using Mask Region-Based Deep Learning Trained on Synthesized Data.

The performance of a neural network depends on the availability of datasets, and most deep learning techniques lack accuracy and generalization when they are trained using limited datasets. Using synthesized training data is one of the effective ways to overcome the above limitation. Besides, the previous corroded bolt detection method has focused on classifying only two classes, clean and fully rusted bolts, and its performance for detecting partially rusted bolts is still questionable. This study presents a deep learning method to identify corroded bolts in steel structures using a mask region-based convolutional neural network (Mask-RCNN) trained on synthesized data. The Resnet50 integrated with a feature pyramid network is used as the backbone for feature extraction in the Mask-RCNN-based corroded bolt detector. A four-step data synthesis procedure is proposed to autonomously generate the training datasets of corroded bolts with different severities. Afterwards, the proposed detector is trained by the synthesized datasets, and its robustness is demonstrated by detecting corroded bolts in a lab-scale steel structure under varying capturing distances and perspectives. The results show that the proposed method has detected corroded bolts well and identified their corrosion levels with the most desired overall accuracy rate = 96.3% for a 1.0 m capturing distance and 97.5% for a 15° perspective angle.

Capsule-Like Smart Aggregate with Pre-Determined Frequency Range for Impedance-Based Stress Monitoring.

In this article, a new capsule-like smart aggregate (CSA) is developed and verified for impedance-based stress monitoring in a pre-determined frequency range of less than 100 kHz. The pros and cons of the existing smart aggregate models are discussed to define the requirement for the improved CSA model. The conceptual design and the impedance measurement model of the capsule-like smart aggregate (CSA) are demonstrated for concrete damage monitoring. In the model, the interaction between the CSA and the monitored structure is considered as the 2-degrees of freedom (2-DOF) impedance system. The mechanical and impedance responses of the CSA are described for two conditions: during concrete strength development and under compressive loadings. Next, the prototype of the CSA is designed for impedance-based monitoring in concrete structures. The local dynamic properties of the CSA are numerically simulated to pre-determine the sensitive frequency bands of the impedance signals. Numerical and experimental impedance analyses are performed to investigate the sensitivity of the CSA under compressive loadings. The changes in the impedance signals of the CSA induced by the compressive loadings are analyzed to assess the effect of loading directions on the performance of the CSA. Correlations between statistical impedance features and compressive stresses are also made to examine the feasibility of the CSA for stress quantification.

Optimal Localization of Smart Aggregate Sensor for Concrete Damage Monitoring in PSC Anchorage Zone.

This study investigates the feasibility of smart aggregate (SA) sensors and their optimal locations for impedance-based damage monitoring in prestressed concrete (PSC) anchorage zones. Firstly, numerical stress analyses are performed on the PSC anchorage zone to determine the location of potential damage that is induced by prestressing forces. Secondly, a simplified impedance model is briefly described for the SA sensor in the anchorage. Thirdly, numerical impedance analyses are performed to explore the sensitivities of a few SA sensors in the anchorage zone under the variation of prestressing forces and under the occurrence of artificial damage events. Finally, a real-scale PSC anchorage zone is experimentally examined to evaluate the optimal localization of the SA sensor for concrete damage detection. Impedance responses measured under a series of prestressing forces are statistically quantified to estimate the performance of damage monitoring via the SA sensor in the PSC anchorage.

Correction: Pham et al. Bolt-Loosening Monitoring Framework Using an Image-Based Deep Learning and Graphical Model. Sensors 2020, 20, 3382.

The authors wish to make the following correction to this paper [...].

A smart LED therapy device with an automatic facial acne vulgaris diagnosis based on deep learning and internet of things application.

In low-level laser therapy, providing an optimal dosage and proposing a proper diagnosis before dermatological treatment are essential to reduce the side effects and potential dangers. In this article, a smart LED therapy system for automatic facial acne vulgaris diagnosis based on deep learning and Internet of Things application is proposed. The main goals of this study were to (1) develop an LED therapy device with different power densities and LED grid control; (2) propose a deep learning model based on modified ResNet50 and YOLOv2 for an automatic acne diagnosis; and (3) develop a smartphone application for facial photography image capture and LED therapy parameter configuration. Furthermore, a healthcare Internet of Things (H-IoT) platform for the connectivity between smartphone apps, the cloud server, and the LED therapy device is proposed to improve the efficiency of the treatment process. Experiments were conducted on test data sets divided by a cross-validation method to verify the feasibility of the proposed LED therapy system with automatic facial acne detection. The obtained results evidenced the practical application of the proposed LED therapy system for automatic acne diagnosis and H-IoT-based solutions.

Monitoring of Corroded and Loosened Bolts in Steel Structures via Deep Learning and Hough Transforms.

In this study, a regional convolutional neural network (RCNN)-based deep learning and Hough line transform (HLT) algorithm are applied to monitor corroded and loosened bolts in steel structures. The monitoring goals are to detect rusted bolts distinguished from non-corroded ones and also to estimate bolt-loosening angles of the identified bolts. The following approaches are performed to achieve the goals. Firstly, a RCNN-based autonomous bolt detection scheme is designed to identify corroded and clean bolts in a captured image. Secondly, a HLT-based image processing algorithm is designed to estimate rotational angles (i.e., bolt-loosening) of cropped bolts. Finally, the accuracy of the proposed framework is experimentally evaluated under various capture distances, perspective distortions, and light intensities. The lab-scale monitoring results indicate that the suggested method accurately acquires rusted bolts for images captured under perspective distortion angles less than 15° and light intensities larger than 63 lux.

Bolt-Loosening Monitoring Framework Using an Image-Based Deep Learning and Graphical Model.

In this study, we investigate a novel idea of using synthetic images of bolts which are generated from a graphical model to train a deep learning model for loosened bolt detection. Firstly, a framework for bolt-loosening detection using image-based deep learning and computer graphics is proposed. Next, the feasibility of the proposed framework is demonstrated through the bolt-loosening monitoring of a lab-scaled bolted joint model. For practicality, the proposed idea is evaluated on the real-scale bolted connections of a historical truss bridge in Danang, Vietnam. The results show that the deep learning model trained by the synthesized images can achieve accurate bolt recognitions and looseness detections. The proposed methodology could help to reduce the time and cost associated with the collection of high-quality training data and further accelerate the applicability of vision-based deep learning models trained on synthetic data in practice.