Publicly accessible weather radar data have significant capabilities for meteorological measurements and predictions and, further, have the potential to measure nonmeteorological events that include smoke, ash, and debris plumes as well as explosions. The ability to identify and track nonmeteorological events can be of assistance in emergency response, hazard mitigation, and related activities in locations where radar coverage both exists and is recorded and accessible to the user. In this study, events from multiple locations in the United States that are reported in news outlets are assessed using a manual inspection process of Level 2 weather radar data to identify anthropogenic and nonbiological returns. Explosive events are also identified, and a large high-altitude debris cloud from the intentional destruction of the SpaceX Starship is tracked across a wide area. Finally, future efforts using a machine learning model are discussed as a means of automating the process and potentially enabling near-real-time nonmeteorological event identification in the same areas where the data are accessible. Using weather radar data can be a valuable new tool for Department of Defense systems to aid in military awareness, and for interagency emergency response and forensic mission experts to consider national weather service data in their mission profiles. Radar data can be effective in detecting several common types of emergencies and inform and aid response personnel.

This study introduces an innovative approach by incorporating statistical offset features, range profiles, time-frequency analyses, and azimuth-range-time characteristics to effectively identify various human daily activities. Our technique utilizes nine feature vectors consisting of six statistical offset features and three principal component analysis network (PCANet) fusion attributes. These statistical offset features are derived from combined elevation and azimuth data, considering their spatial angle relationships. The fusion attributes are generated through concurrent 1D networks using CNN-BiLSTM. The process begins with the temporal fusion of 3D range-azimuth-time data, followed by PCANet integration. Subsequently, a conventional classification model is employed to categorize a range of actions. Our methodology was tested with 21,000 samples across fourteen categories of human daily activities, demonstrating the effectiveness of our proposed solution. The experimental outcomes highlight the superior robustness of our method, particularly when using the Margenau-Hill Spectrogram for time-frequency analysis. When employing a random forest classifier, our approach outperformed other classifiers in terms of classification efficacy, achieving an average sensitivity, precision, F1, specificity, and accuracy of 98.25%, 98.25%, 98.25%, 99.87%, and 99.75%, respectively.

Falls among the elderly are a common and serious health risk that can lead to physical injuries and other complications. To promptly detect and respond to fall events, radar-based fall detection systems have gained widespread attention. In this paper, a deep learning model is proposed based on the frequency spectrum of radar signals, called the convolutional bidirectional long short-term memory (CB-LSTM) model. The introduction of the CB-LSTM model enables the fall detection system to capture both temporal sequential and spatial features simultaneously, thereby enhancing the accuracy and reliability of the detection. Extensive comparison experiments demonstrate that our model achieves an accuracy of 98.83% in detecting falls, surpassing other relevant methods currently available. In summary, this study provides effective technical support using the frequency spectrum and deep learning methods to monitor falls among the elderly through the design and experimental validation of a radar-based fall detection system, which has great potential for improving quality of life for the elderly and providing timely rescue measures.

To address the challenges in recognizing various types of falls, which often exhibit high similarity and are difficult to distinguish, this paper proposes a human fall classification system based on the SE-Residual Concatenate Network (SE-RCNet) with adaptive weighted fusion. First, we designed the innovative SE-RCNet network, incorporating SE modules after dense and residual connections to automatically recalibrate feature channel weights and suppress irrelevant features. Subsequently, this network was used to train and classify three types of radar images: time-distance images, time-distance images, and distance-distance images. By adaptively fusing the classification results of these three types of radar images, we achieved higher action recognition accuracy. Experimental results indicate that SE-RCNet achieved F1-scores of 94.0%, 94.3%, and 95.4% for the three radar image types on our self-built dataset. After applying the adaptive weighted fusion method, the F1-score further improved to 98.1%.

Objective: This study evaluates the agreement between a new low-load sleep monitoring system, QSA600, based on millimeter-wave radar technology, and polysomnography (PSG) in diagnosing obstructive sleep apnea (OSA). Methods: A total of 155 subjects were recruited for a parallel agreement study in the sleep laboratory of the Department of Otorhinolaryngology Head and Neck Surgery at Shanghai Sixth People\'s Hospital from July to September 2023. The subjects underwent simultaneous monitoring with both PSG and the QSA600 system. One hundred and forty-five subjects consisting of 75 males and 70 females included in the final analysis, with an average age of (35.30±12.41) years, an average height of (168.23±8.08) cm, and an average weight of (68.28±13.74) kg. The subjects were divided into four groups based on the apnea-hypopnea index (AHI): <5.0 events/h (non-OSA group, 39 cases), ≥5.0-<15.0 events/h (mild OSA group, 47 cases), ≥15.0-<30.0 events/h (moderate OSA group, 25 cases), and≥30.0 events/h (severe OSA group, 34 cases). Intraclass correlation coefficients (ICC), Pearson correlation coefficients (r), and Bland-Altman analysis were employed to assess the agreement between the two monitoring techniques regarding AHI and other parameters. Sensitivity and specificity of the QSA600 in diagnosing OSA were evaluated at different AHI thresholds. Statistical analyses were conducted using MATLAB R2022a. Results: Using AHI 5 events/h, 15 events/h and 30 events/h as thresholds, the sensitivity for diagnosing mild, moderate, and severe OSA was 88.68%, 89.83% and 97.06%, respectively. The specificity was 94.87%, 98.84% and 99.10%, respectively. The areas under the receiver operating characteristic (ROC) curve was 0.973 4, 0.990 9 and 0.999 5, respectively. The comparison of key indicators between QSA600 and PSG diagnostic results revealed:a Pearson correlation coefficient of 0.987 2(P<0.001) between the AHI measurement values. The mean difference between the Bland-Altman measurement values of the two was -1.43(95%CI:-8.74-5.88) events/h and the ICC between the two was 0.985 0(95%CI: 0.975 4-0.990 4). Conclusions: As a new low-load sleep monitoring system, QSA600 demonstrates high concordance with traditional PSG in diagnosing OSA and stratifying its severity, which has promising potential for clinical application. (Clinical trial registration number: NCT06038006).
目的： 评价基于毫米波雷达的低负荷新型睡眠监测系统QSA600与多导睡眠监测（polysomnography，PSG）在诊断阻塞性睡眠呼吸暂停（obstructive sleep apnea，OSA）的一致性。 方法： 2023年7月至9月招募155名受试者，在上海市第六人民医院耳鼻咽喉头颈外科睡眠实验室同时接受PSG和QSA600监测，145名受试者纳入最终分析，其中男75例，女70例，年龄为（35.30±12.41）岁，身高为（168.23±8.08）cm，体重为（68.28±13.74）kg，进行一致性平行对照研究。以呼吸暂停低通气指数（apnea hypopnea index，AHI）将受试者分为4组：<5.0 次/h（NOSA组，39例），≥5.0次/h~<15.0 次/h（轻度OSA组，47例），≥15.0次/h~<30.0 次/h（中度OSA组，25例），≥30.0 次/h（重度OSA组，34例）。采用组内相关系数（intraclass correlation coefficients，ICC）、Pearson相关系数（r）和Bland-Altman检验来评价2种监测技术AHI等指标的一致性。根据AHI不同阈值，检验QSA600的诊断灵敏度、特异度。采用MATLAB R2022a进行统计学分析。 结果： 分别以AHI 5 次/h、15 次/h、30 次/h为阈值，对轻度、中度、重度OSA诊断的灵敏度分别为88.68%、89.83%、97.06%，特异度分别为94.87%、98.84%、99.10%，受试者工作特征（receiver operating characteristic，ROC）曲线下面积分别为0.973 4、0.990 9、0.999 5。QSA600与PSG诊断结果的关键指标（AHI）的一致性对比结果：二者Pearson相关性为0.987 2（P<0.001）；二者Bland-Altman测定差值均数为-1.43次/h［95%置信区间（confidence interval，CI）为-8.74~5.88］；二者ICC为0.985 0（95%CI为0.975 4~0.990 4）。 结论： 作为低负荷新型睡眠监测系统，QSA600在诊断OSA及不同严重程度OSA分级方面与PSG表现出很高的一致性，有较好临床应用前景。.

Radio detection and ranging-based (radar) sensing offers unique opportunities for biomedical monitoring and can help overcome the limitations of currently established solutions. Due to its contactless and unobtrusive measurement principle, it can facilitate the longitudinal recording of human physiology and can help to bridge the gap from laboratory to real-world assessments. However, radar sensors typically yield complex and multidimensional data that are hard to interpret without domain expertise. Machine learning (ML) algorithms can be trained to extract meaningful information from radar data for medical experts, enhancing not only diagnostic capabilities but also contributing to advancements in disease prevention and treatment. However, until now, the two aspects of radar-based data acquisition and ML-based data processing have mostly been addressed individually and not as part of a holistic and end-to-end data analysis pipeline. For this reason, we present a tutorial on radar-based ML applications for biomedical monitoring that equally emphasizes both dimensions. We highlight the fundamentals of radar and ML theory, data acquisition and representation and outline categories of clinical relevance. Since the contactless and unobtrusive nature of radar-based sensing also raises novel ethical concerns regarding biomedical monitoring, we additionally present a discussion that carefully addresses the ethical aspects of this novel technology, particularly regarding data privacy, ownership, and potential biases in ML algorithms.

Current methods for fall risk assessment rely on Quantitative Gait Analysis (QGA) using costly optical tracking systems, which are often only available at specialized laboratories that may not be easily accessible to rural communities. Radar placed in a home or assisted living facility can acquire continuous ambulatory recordings over extended durations of a subject\'s natural gait and activity. Thus, radar-based QGA has the potential to capture day-to-day variations in gait, is time efficient and removes the burden for the subject to come to a clinic, providing a more realistic picture of older adults\' mobility. Although there has been research on gait-related health monitoring, most of this work focuses on classification-based methods, while only a few consider gait parameter estimation. On the one hand, metrics that are accurately and easily computable from radar data have not been demonstrated to have an established correlation with fall risk or other medical conditions; on the other hand, the accuracy of radar-based estimates of gait parameters that are well-accepted by the medical community as indicators of fall risk have not been adequately validated. This paper provides an overview of emerging radar-based techniques for gait parameter estimation, especially with emphasis on those relevant to fall risk. A pilot study that compares the accuracy of estimating gait parameters from different radar data representations - in particular, the micro-Doppler signature and skeletal point estimates - is conducted based on validation against an 8-camera, marker-based optical tracking system. The results of pilot study are discussed to assess the current state-of-the-art in radar-based QGA and potential directions for future research that can improve radar-based gait parameter estimation accuracy.

Falls among the elderly population pose significant health risks, often leading to morbidity and decreased quality of life. Traditional fall detection methods, namely wearable devices and cameras, have limitations such as lighting conditions and privacy concerns. Radar-based fall detection has emerged as a promising alternative, offering unobtrusive technique. In this study, an attempt has been made to classify fall detection using smoothed pseudo wigner-ville distribution (SPWVD) images and XGBoost learning. For this, online publicly available radar database (N=15) is considered. Radar signals is employed to SPWVD for time-frequency representation images. Ten features are extracted and applied to XGBoost learning. Experiments are performed and performance is evaluated using 10-fold cross validation. The proposed approach is able to discriminate elderly fall. Using XGBoost learning, the approach yields a maximum average classification accuracy, f1-score, precision, sensitivity, specificity, and kappa scores of 87.47%, 87.38%, 88.12%, 86.81%, 88.31% and 74.94% respectively. The combination of conventional features with concentration measures and median frequency obtained the second best performance. Thus, the proposed framework could be utilized for accurate and efficient detection of falls among the elderly population in their private spaces.

BACKGROUND: Changes in spatio-temporal gait parameters and their variability during balance-challenging tasks are markers of motor performance linked to fall risk. Radio frequency (RF) sensors hold great promise towards achieving continuous remote monitoring of these parameters.
OBJECTIVE: To establish the concurrent validity of RF-based gait metrics extracted using micro-Doppler (µD) signatures and to determine whether these metrics are sensitive to gait modifications created by multidirectional visual perturbations.
METHODS: Fifteen participants walked overground in a virtual environment (VE) and VE with medio-lateral (ML) and antero-posterior (AP) perturbations. An optoelectronic motion capture system and one RF sensor were used to extract the linear velocity of the trunk and estimate step time (ST), step velocity (SV), step length (SL), and their variability (STV, SVV, and SLV). Intra-class coefficient for consistency (ICC), mean and standard deviation of the differences (MD), 95 % limits of agreement, and Pearson correlation coefficients (r) were used to determine concurrent validity. One-way repeated-measures analysis of variance was used to analyze the main and interaction effects of visual conditions.
RESULTS: All outcomes showed good to excellent reliability (r>0.795, ICC>0.886). Average gait parameters showed good to excellent agreement, with values obtained with the RF sensor systematically smaller than the values obtained with the markers (MD of 0.001 s, 0.09 m/s, and 0.06 m). Gait variability parameters showed poor to moderate agreement, with values obtained with the RF sensor systematically larger than those obtained with the markers (MD of 1.9 %-3.9 %). Both measurement systems reported decreased SL and SV during ML perturbations, but the gait variability parameters extracted with the radar were not able to detect the higher STV and SLV during this condition.
CONCLUSIONS: The radar µD signature is a valid and reliable method for the assessment of average spatio-temporal gait parameters but gait variability measures need to be viewed with caution because of the lower levels of agreement and sensitivity to ML visual perturbations. This work represents an initial investigation for the development of a low-cost system that will facilitate aging-in-place by providing remote monitoring of gait in natural settings.

The subsidence of the earth\'s surface in mining areas is characterized by fast speed and large gradients. Conventional small baseline subset interferometric synthetic aperture radar (SBAS-InSAR) monitoring can significantly underestimate results, making it challenging to capture the surface\'s temporal subsidence features. In this context, this paper proposes a method for monitoring subsidence in mining areas. It utilizes a phase unwrapping network (PUNet) and a fused Weibull model within the SBAS-InSAR framework to address nonlinear and large-gradient subsidence. The basic principle of this method is to first process the SAR images using the small baseline method to obtain the differential interferogram, utilizing the PUNet to obtain reliable large-gradient unwrapped phases. Next, the Weibull model parameters of each pixel are calculated based on the unwrapped phase, and the temporal subsidence of each point on the surface is determined using the calculated parameters. This method introduces a nonlinear model into the SBAS-InSAR solution, which is more consistent with the subsidence characteristics of mining areas. Through experimentation in a backfilled mining working face, the proposed method in this paper yields superior monitoring results compared to conventional approaches.