Explore the vast range of titles available.

Purpose In craniospinal irradiation, two or three isocenter groups along the craniocaudal axis are required to cover the long treatment target. Adapting the isocenter distance according to daily deviations in patient position is challenging because dosimetric hot or cold spots may occur in the field junction. The aim of this study was to quantify the effect of adapting the isocenter distance to patient position on the dose distribution of the field overlap region in craniospinal irradiation using partial‐arc volumetric modulated arc therapy. Methods The magnitude of isocenter distance deviations in craniocaudal direction was quantified by registering the setup images of 204 fractions of 12 patients to the planning images. The dosimetric effect of these deviations was determined by shifting the isocenters of the original treatment plan and calculating the resulting dose distribution. Results On fraction‐level, deviations larger than 3 mm caused more than 5 percentage point changes in the doses covering 2% (D2%) and 98% (D98%) of the junction volume in several patients. On treatment course‐level, the changes in D2% and D98% of the junction volume were less than 5 percentage points in all cases except for one patient. Conclusions Craniocaudal isocenter distance adaptation can be conducted provided that the mean isocenter distance deviation over the treatment course is within 3 mm.

In this paper, we consider a new design problem of optimizing a linear transceiver for correlated multiple-input multiple-output (MIMO) interference channels in the presence of channel state information (CSI) errors, which is a more realistic and practical scenario than those considered in the previous studies on uncorrelated MIMO interference channels. By taking CSI errors into account, the optimization problem is initially formulated to minimize the average mean square error (MSE) under the general power constraints. Since the objective function is not jointly convex in precoders and receive filters, we split the original problem into two convex subproblems, and then linear precoders and receive filters are obtained by solving two subproblems iteratively. It is shown that the proposed algorithm is guaranteed to converge to a local minimum. The numerical results show that the proposed algorithm can significantly reduce the sensitivity to CSI errors compared with the existing robust schemes in the correlated MIMO interference channel.

Background Cardiovascular phosphorus MR spectroscopy ( 31 P-CMRS) is a powerful tool for probing energetics in the human heart, through quantification of phosphocreatine (PCr) to adenosine triphosphate (ATP) ratio. In principle, 31 P-CMRS can also measure cardiac intracellular pH (pH i ) and the free energy of ATP hydrolysis (ΔG ATP ). However, these require determination of the inorganic phosphate (Pi) signal frequency and amplitude that are currently not robustly accessible because blood signals often obscure the Pi resonance. Typical cardiac 31 P-CMRS protocols use low (e.g. 30°) flip-angles and short repetition time (TR) to maximise signal-to-noise ratio (SNR) within hardware limits. Unfortunately, this causes saturation of Pi with negligible saturation of the flowing blood pool. We aimed to show that an adiabatic 90° excitation, long-TR, 7T 31 P-CMRS protocol will reverse this balance, allowing robust cardiac pH i measurements in healthy subjects and patients with hypertrophic cardiomyopathy (HCM). Methods The cardiac Pi T 1 was first measured by the dual TR technique in seven healthy subjects. Next, ten healthy subjects and three HCM patients were scanned with 7T 31 P-MRS using long (6 s) TR protocol and adiabatic excitation. Spectra were fitted for cardiac metabolites including Pi. Results The measured Pi T 1 was 5.0 ± 0.3 s in myocardium and 6.4 ± 0.6 s in skeletal muscle. Myocardial pH was 7.12 ± 0.04 and Pi/PCr ratio was 0.11 ± 0.02. The coefficients of repeatability were 0.052 for pH and 0.027 for Pi/PCr quantification. The pH in HCM patients did not differ ( p  = 0.508) from volunteers. However, Pi/PCr was higher (0.24 ± 0.09 vs. 0.11 ± 0.02; p  = 0.001); Pi/ATP was higher (0.44 ± 0.14 vs. 0.24 ± 0.05; p  = 0.002); and PCr/ATP was lower (1.78 ± 0.07 vs. 2.10 ± 0.20; p  = 0.020), in HCM patients, which is in agreement with previous reports. Conclusion A 7T 31 P-CMRS protocol with adiabatic 90° excitation and long (6 s) TR gives sufficient SNR for Pi and low enough blood signal to permit robust quantification of cardiac Pi and hence pH i . Pi was detectable in every subject scanned for this study, both in healthy subjects and HCM patients. Cardiac pH i was unchanged in HCM patients, but both Pi/PCr and Pi/ATP increased that indicate an energetic impairment in HCM. This work provides a robust technique to quantify cardiac Pi and pH i .

This paper studies a novel movable antenna (MA)-aided Cell-Free Massive MIMO system to leverage the corresponding spatial degrees of freedom (DoFs) for improving the performance of distributed wireless networks. We aim to maximize the ergodic sum capacity by jointly optimizing the MA positions and the transmit covariance matrix based on statistical channel state information (CSI). To address the non-convex stochastic optimization problem, we propose a novel Constrained Stochastic Successive Convex Approximation (CSSCA) framework, enhanced with a robust slack-variable mechanism to handle non-convex antenna spacing constraints and ensure iterative feasibility. Numerical results show that the considered MA-enhanced system can significantly improve the ergodic capacity compared to fixed-antenna cell-free systems and that the proposed algorithm exhibits robust convergence behavior.

This paper studies the energy efficiency (EE) oriented precoding design in multi-cell massive multiple-input multiple-output (MIMO) systems, with only statistical channel state information (CSI) at the transmitter. During the transmission, as the channel varies dynamically with time and the previously obtained CSI becomes outdated, the base stations must adjust their transmit policies accordingly. To tackle this issue, we propose an online EE maximization algorithm that can achieve a no-regret transmission; i.e., the performance of this online method gradually approaches that of the fixed offline method which has full knowledge of the future CSI. Specifically, we first construct the online EE optimization problem in a distributed way to reduce the information required to be exchanged between cells. Then, we apply the large-dimensional random matrix theory to lower the calculation complexity, and the Charnes-Cooper transform to address the nonconvexity of the problem, respectively. The online gradient ascent method is utilized to perform this no-regret power allocation strategy based on all past CSI. We also assess the robustness of the algorithm to estimation error of statistical CSI under some mild conditions which can usually be satisfied in practice. Numerical results demonstrate the no-regret property and the robustness of the proposed online algorithm for energy efficient multi-cell massive MIMO transmission.

The rapid development of machine learning and wireless communication is creating a new paradigm for future networks, namely edge-intelligent networks. Specifically, data generated by terminal devices is processed via machine learning at the edge of wireless networks, but not at the cloud. Owing to the growing concern for privacy information sharing, federated learning, as a new branch of machine learning, is appealing in edge-intelligent networks. For federated learning, the wireless transmission capabilities under practical conditions, e.g., imperfect channel state information (CSI), have a great impact on the accuracy of global aggregation of local model updates. Therefore, it is very important to enhance the robustness of communication for federated learning. In order to realize robust communication in the presence of channel uncertainty, we propose a robust federated learning algorithm for edge-intelligent networks, including device selection, transmit power allocation, and receive beamforming. Simulation results validate the robustness and effectiveness of the proposed robust federated learning algorithm in edge-intelligent networks.

PurposeTo investigate the dose stability of craniospinal irradiation based on irradiation method of proton beam therapy (PBT).Methods and materialsTwenty-four pediatric and young adult brain tumor patients (age: 1–24 years) were examined. Treatment method was passive-scattered PBT (PSPT) in 8 patients and intensity-modulated PBT (IMPT) in 16 patients. The whole vertebral body (WVB) technique was used in 13 patients whose ages were younger than 10, and vertebral body sparing (VBS) technique was used for the remaining 11 patients aged 10 and above. Dose stability of planning target volume (PTV) against set-up error was investigated.ResultsThe minimum dose (Dmin) of IMPT was higher than that of PSPT (p = 0.01). Inhomogeneity index (INH) of IMPT was lower than that of PSPT (p = 0.004). When the irradiation field of the cervical spinal cord level (C level) was shifted, the maximum dose (Dmax) was lower in IMPT, and mean dose (Dmean) was higher than PSPT as movement became greater to the cranial–caudal direction (p = 0.000–0.043). Dmin was higher and INH was lower in IMPT in all directions (p = 0.000–0.034). When the irradiation field of the lumber spinal cord level (L level) was shifted, Dmax was lower in IMPT as movement became greater to the cranial direction (p = 0.000–0.028). Dmin was higher and INH was lower in IMPT in all directions (p = 0.000–0.022).ConclusionsThe PTV doses of IMPT and PSPT are robust and stable in both anterior–posterior and lateral directions at both C level and L level, but IMPT is more robust and stable than PSPT for cranial–caudal movements.Trial registryClinical Trial Registration number: No. 04-03.

This work puts the first effort on investigating robust and adaptive pedestrian passing detection and direction recognition based on WiFi Channel State Information (CSI). Specifically, we first give an insight into the challenges as well as opportunities of realizing cross-scenario pedestrian monitoring based on comprehensive analysis of CSI patterns. In light of the findings, we design a novel system framework, which consists of an offline pattern clustering and training module for constructing a unified offline database, and an online adaptive monitoring module for enabling real-time pedestrian passing detection and direction recognition. Further, we propose four mechanisms, including a CSI pre-processing method to enhance system robustness by extracting stable and distinct CSI features, a Two-stage Clustering (TC) method to enable cross-scenario CSI feature classification by segmenting the offline datasets automatically, a unified segmenting and detecting (USD) method to enable adaptive pedestrian passing detection by training a component classifier and a sample classifier, and a dynamic direction calculation (DDC) method to recognize the passing direction based on time of passing estimation, link confidence calculation, and direction indicator calculation. Finally, we implement the system prototype and evaluate the system performance in real-world scenarios. A comprehensive experimental study demonstrates that the proposed framework and the mechanisms can effectively enhance system robustness and adaptiveness on pedestrian monitoring.

Human activity recognition (HAR) is crucial in applications such as smart homes, interactive games, surveillance, security, and healthcare. In recent years, Channel State Information (CSI) data extracted from Wi-Fi signals has garnered significant interest for applications in HAR. This interest stems from CSI’s several advantages, including its immunity to illumination variations and environmental disturbances, and the elimination of the need for wearable devices. Despite being widely used, existing HAR system’s performance suffers when used in new environments without system improvement or retraining. This constraint can be overcome by gathering and annotating data from various locations, and then re-training the system. However, it is far from ideal from the privacy perspective, as the training algorithms access the data from different privacy-sensitive environments. This motivates us to design a reliable and robust privacy-preserving HAR system. In this work, we introduce a Differentially Private Principal Component-based Wavelet Convolutional Neural Network (DP-PCWCNN) that offers accurate and robust HAR performance across different environments, while preserving strict privacy constraints. We evaluate the performance of our proposed algorithm on two publicly available real datasets and demonstrate that our proposed system closely approximates the non-private system’s performance for some parameter choices.

Reconfigurable Intelligent Surface (RIS) has emerged as a promising enabling technology for wireless communications, which significantly enhances system performance through real-time manipulation of electromagnetic wave reflection characteristics. In RIS-assisted communication systems, existing deep learning-based channel state information (CSI) feedback methods often suffer from excessive parameter requirements and high computational complexity. To address this challenge, this paper proposes LwCSI-Net, a lightweight autoencoder network specifically designed for RIS-assisted multiple-input single-output (MISO) systems, aiming to achieve efficient and low-complexity CSI feedback. The core contribution of this work lies in an innovative lightweight feedback architecture that deeply integrates multi-layer convolutional neural networks (CNNs) with attention mechanisms. Specifically, the network employs 1D convolutional operations with unidirectional kernel sliding, which effectively reduces trainable parameters while maintaining robust feature-extraction capabilities. Furthermore, by incorporating an efficient channel attention (ECA) mechanism, the model dynamically allocates weights to different feature channels, thereby enhancing the capture of critical features. This approach not only improves network representational efficiency but also reduces redundant computations, leading to optimized computational complexity. Additionally, the proposed cross-channel residual block (CRBlock) establishes inter-channel information-exchange paths, strengthening feature fusion and ensuring outstanding stability and robustness under high compression ratio (CR) conditions. Our experimental results show that for CRs of 16, 32, and 64, LwCSI-Net significantly improves CSI reconstruction performance while maintaining fewer parameters and lower computational complexity, achieving an average complexity reduction of 35.63% compared to state-of-the-art (SOTA) CSI feedback autoencoder architectures.