As a convenient and natural way of human-computer interaction, gesture recognition technology has broad research and application prospects in many fields, such as intelligent perception and virtual reality. This paper summarized the relevant literature on gesture recognition using Frequency Modulated Continuous Wave (FMCW) millimeter-wave radar from January 2015 to June 2023. In the manuscript, the widely used methods involved in data acquisition, data processing, and classification in gesture recognition were systematically investigated. This paper counts the information related to FMCW millimeter wave radar, gestures, data sets, and the methods and results in feature extraction and classification. Based on the statistical data, we provided analysis and recommendations for other researchers. Key issues in the studies of current gesture recognition, including feature fusion, classification algorithms, and generalization, were summarized and discussed. Finally, this paper discussed the incapability of the current gesture recognition technologies in complex practical scenes and their real-time performance for future development.
Extremely low frequency (ELF) magnetic field (MF) exposure in electric vehicles (EVs) has raised public concern for human health. There have been many studies evaluating magnetic field values in these vehicles. However, there has been no report on the temporal variation of the magnetic field in the cabin. This is the first study on the long-term monitoring of actual MFs in EVs. In the study, we measured the magnetic flux density (B) in three shared vehicles over a period of two years. The measurements were performed at the front and rear seats during acceleration and constant-speed driving modes. We found that the B amplitudes and the spectral components could be modified by replacing the components and the hubs, while regular checks or maintenance did not influence the B values in the vehicle. This observation highlights the necessity of regularly monitoring ELF MF in EVs, especially after major repairs or accidents, to protect car users from potentially excessive ELF MF exposure. These results should be considered in updates of the measurement standards. The ELF MF effect should also be taken into consideration in relevant epidemiological studies.
International guidelines and standards for human protection from electromagnetic fields have been revised recently. The epithelial/absorbed power density (APD) has been used as a new physical quantity for local exposure to frequencies > 6 GHz, related to the temperature rise in the superficial layer. The assessment methods of APD for practical exposure scenarios are crucial for realistic exposures. This study investigates averaging methods for APD in nonplanar skin models for local electromagnetic field exposure from 10 to 90 GHz. Eight research groups compared the calculated APDs using distinct numerical approaches and postprocessing techniques. The intercomparison aimed to clarify the primary causes of variations in aspects, such as APD averaging methods, skin model structure, and types of exposure antennas. Statistical analyses reveal that the maximum differences in relative confidence intervals (RCI) due to different average methods and skin models are within 9.4% and 5.1%, respectively. In contrast, when the distance between the antenna and skin is set to 5 mm, the discrepancy attributed to the exposure antenna reaches 59.4% at 10 GHz. This difference does not exceed 9.2% under other computational conditions. Additionally, the spatially averaged APD appears to have a linear relationship with the maximum skin surface temperature elevation, based on regression analysis. The findings indicate that the variances in the spatially averaged APD are largely independent of both the APD averaging methods and skin model structures. However, they slightly depend on the antenna types used as exposure sources.
Image segmentation plays a vital role in the medical diagnosis and intervention field. The segmentation methods can be classified as fully automated, semiautomated or manual. Among them, manual segmentation can best improve the quality of the results, but it is time-consuming and tedious, and it may lead to operator bias. A continuity-aware probabilistic network based on the divide-and-conquer method was proposed in the current work. The proposed network comprised backbone network, local segmentation and a weight network. The backbone network extracts the features from image. The local segmentation divides the data space, whereas the weight network provides the continuity-aware weights. Therefore, combining those results of the weighted segments can eventually yield precise estimations. In this study, the proposed model was evaluated against several recent methods on the three datasets, and a several performance indexes of segmentation were evaluated for liver segmentation, the results showing that it is the most advanced liver segmentation approach. The source code of this work is publicly shared at https://github.com/licongsheng/DCSegNet for others to easily reproduce the work and build their own models with the introduced mechanisms.
Liver and liver tumor segmentation based on abdomen computed tomography (CT) images is an essential step in computer-assisted clinical interventions. However, liver and tumor segmentation remains the difficult issue in the medical image processing field, which is ascribed to the anatomical complexity of the liver and the poor demarcation between the liver and other nearby organs on the image. The existing 3D automatic liver and tumor segmentation algorithms based on full convolutional networks, such as V-net, have utilized the loss functions on the basis of integration (summing) over a segmented region (like Dice or cross-entropy). Unfortunately, the number of foreground and background voxels is usually highly imbalanced in liver and tumor segmentation tasks. This greatly varies the value of regional loss between various segmentation classes, and affects the training stability and effect. In the present study, an improved V-net algorithm was applied for 3D liver and tumor segmentation based on region and distance metrics. The distance metric-based loss function utilized a distance metric of the contour (or shape) space rather than the area. The model was jointly trained by the original regional loss and the three distance-based loss functions (including Boundary (BD) loss, Hausdorff (HD) loss, and Signed Distance Map (SDM) loss) to solve the problem of the highly unbalanced liver and tumor segmentation. Besides, the algorithm was tested in two databases LiTS 2017 (Technical University of Munich, Munich, Germany, 2017) and 3D-IRCADb (Research Institute against Digestive Cancer, Strasbourg Cedex, France, 2009), and the results proved the effectiveness of improvement.
Anatomical human models have been widely used in the assessment of induced field strength for low-frequency (LF) electromagnetic field exposure. One bottleneck is the assignment of a single electrical conductivity to all the voxels of the corresponding tissue. This simplification is known to cause computational artifact; therefore, a large reduction factor was considered in international guidelines and standards. Recently, head models with nonuniform conductivities generated using deep learning networks were proposed, and the effect on the reduction of staircasing artifacts was demonstrated. If the effectiveness of the models is confirmed for different models and codes, it would be useful to derive the relationship between the internal and external field strengths needed for setting the exposure limit. The Subcommittee 6 of the IEEE International Committee on Electromagnetic Safety Technical Committee 95 launched a working group to conduct the first intercomparison study of the induced electric field in learning-based head models exposed to LF magnetic fields. Seven international research groups have cooperated in this joint study. The highest relative difference (RD) in averaged electric fields was 23%, which is attributable to the difference caused the by scalar potential finite difference (SPFD) method and finite element method. Except for one group, the RDs in the 100 th and 99 th percentile values of the averaged electric field using the SPFD method with different solvers and codes were below 1%, indicating that the uncertainty due to different codes is sufficiently small under the same exposure scenarios. The findings would be informative for future revision of exposure limits and reduction factors in the exposure standard, which is closely related to computational uncertainty.