Multimodal-attributed graphs (MAGs) couple graph topology with node semantics from text, images, and other modalities. Traditional graph learning contextualizes node semantics by coupling topology with node features. However, this coupling design becomes troublesome in MAGs, where structure-induced and modality-intrinsic semantics may contribute differently to downstream tasks. Structure-induced semantics promote relational consistency through smooth topological variation, whereas modality-intrinsic semantics often encode local, fine-grained distinctions that should not be uniformly smoothed or aligned. Therefore, the key challenge is to identify semantic roles before cross-modal fusion. To this end, we leverage graph-frequency variation as a prior, where low-frequency components capture topology-consistent semantics and high-frequency components preserve modality-specific semantics. Based on this intuition, we propose SMGFM, a spectral multimodal graph pretraining framework that decomposes each modality-specific node signal into graph-frequency bands and assigns band-level semantic roles before cross-modal interaction. Concretely, SMGFM constructs frequency-resolved modality tokens with scalable Chebyshev filters, estimates their coupling reliability through topology-conditioned routing, and performs band-modality interaction before fusion. Its frequency-routed objectives align smooth consensus routes while preserving modality-specific routes, mitigating spatial-domain entanglement and uniform cross-modal alignment. Extensive experiments conducted on the MAG datasets demonstrate that SMGFM achieves state-of-the-art performance across graph-level and modality-level tasks.

Causal discovery seeks to uncover the causal dependencies among variables. For this purpose, we propose an algorithm called Tensor-based Second-order Causal Discovery (TSCD). Its input is a tensor obtained from the covariance matrices of observational and interventional data. Assuming the causal dependencies follow a linear structural equation model on a directed acyclic graph (DAG), TSCD outputs the DAG and the functions on its edges, requiring only that the noise variables are uncorrelated. We also implement a version of the approach for nonlinear models. Our focus on second-order statistics (via the covariance matrices) is motivated by their statistical and computational efficiency relative to higher-order moments, their identifiability relative to first-order statistics, and that they work regardless of whether the variables are Gaussian. We show that TSCD has identifiable causal order and parameters from a number of interventions that is logarithmic in the number of variables. Experiments show that TSCD is robust to noise, competitive with existing methods, and scales to hundreds of variables.

Melanoma is the most dangerous form of skin cancer with five-year survival rates exceeding 99% when detected early but falling sharply once the disease spreads. This paper proposes and evaluates a two-stage fine-tuning approach for ResNet50 applied to binary melanoma classification on dermoscopic images. The core challenges addressed are class imbalance and suboptimal transfer learning from single-stage fine-tuning. After stratified train/validation/test splitting, random oversampling was applied exclusively to the training set to achieve a 1:1 class balance. Stage 1 trained only the classification head with the ResNet50 base frozen, while Stage 2 fine-tuned all layers jointly at a low learning rate of 1e-5 to prevent catastrophic forgetting of learned visual features. On an independent test set of 3,826 images, the model achieved an AUC-ROC of 0.9559, accuracy of 88.34%, sensitivity of 87.56%, specificity of 89.13%, and F1-score of 88.29%. An ablation study confirms the two-stage protocol significantly outperforms single-stage fine-tuning, with sensitivity gains of over 4%. Grad-CAM visualizations demonstrate correct lesion localization. A fully deployable Streamlit detection application is provided alongside all training code.

Personalized medicine in acute ischemic stroke requires moving beyond average treatment effects (ATE) to individualized treatment effect (ITE) estimates to support treatment decisions. In acute ischemic stroke, mechanical thrombectomy has been shown to be more effective on average than lysis in randomized controlled trials (RCTs), such as the MR CLEAN study. We aim to identify which individual patients benefit most from mechanical thrombectomy compared to lysis. The outcome of interest is the modified Rankin Scale (mRS) at three months, an ordinal measure of functional disability (0: no symptoms, 6: death). We demonstrate that causal transformation models on directed acyclic graphs (TRAM-DAG) can be used for ITE estimation after being fitted on observational MAGIC multi-center stroke patient data. To ensure comparability with the MR CLEAN population, which we use for validation, we train the TRAM-DAG on a MAGIC sub-population with NIHSS at admission >= 6, corresponding to one inclusion criterion of MR CLEAN. The fitted model is then used to estimate ITEs for stroke patients in the MR CLEAN population. While these ITE estimates cannot be confirmed experimentally, we show that their average is consistent with the trial's reported ATE. Furthermore, the ITE estimates correctly rank trial patients by their observed frequency of a good outcome (mRS at three months <= 2). These findings support the use of causal models like TRAM-DAG for personalized decision-making in stroke care and highlight their ability to bridge the gap between observational evidence and clinical trials.

Screenshot-based mobile GUI agents can operate ordinary smartphone apps through the same visual interface as a human user, but this capability also turns every screen observation into a privacy boundary. During normal task execution, screenshots may expose contacts, messages, photos, files, recommendations, health cues, and other sensitive context that is unrelated to the user's request. We call this problem incidental visual privacy exposure. It is difficult to address with existing defenses: text anonymization misses many visual and inferential cues, while generic privacy masking can remove the evidence and controls that a GUI agent needs to complete the task. This paper presents CAPED, a context-aware pre-upload exposure control layer for mobile GUI agents. CAPED is designed as a phone-side protection layer: before screenshots are released to a remote multimodal agent, it extracts task requirements, uses screen context as a privacy prior, parses visible UI elements, and selectively exposes only content needed for the current task while masking incidental private content. We evaluate CAPED on AndroidWorld for broad task utility and with a controlled 28-task seeded privacy evaluation used as a measurement instrument for trajectory-level incidental leakage. In this seeded evaluation, Full CAPED reduces success-conditioned weighted seeded leakage from 0.766 under raw screenshots to 0.268 while preserving high task utility. A broader AndroidWorld run shows a remaining prototype-level utility cost, but the results show that task-driven selective exposure can reduce incidental visual leakage before screenshots are released to a remote GUI agent.

For dementia screening and monitoring, standardized tests play a key role in clinical routine since they aim at minimizing subjectivity by measuring performance on a variety of cognitive tasks. In this paper, we report on a study that consists of a semi-standardized history taking followed by two standardized neuropsychological tests, namely the SKT and the CERAD-NB. The tests include basic tasks such as naming objects, learning word lists, but also widely used tools such as the MMSE. Most of the tasks are performed verbally and should thus be suitable for automated scoring based on transcripts. For the first batch of 30 patients, we analyze the correlation between expert manual evaluations and automatic evaluations based on manual and automatic transcriptions. For both SKT and CERAD-NB, we observe high to perfect correlations using manual transcripts; for certain tasks with lower correlation, the automatic scoring is stricter than the human reference since it is limited to the audio. Using automatic transcriptions, correlations drop as expected and are related to recognition accuracy; however, we still observe high correlations of up to 0.98 (SKT) and 0.85 (CERAD-NB). We show that using word alternatives helps to mitigate recognition errors and subsequently improves correlation with expert scores.

In difference-in-differences designs, the parallel trends assumption requires that the outcome gap between treated and control units would have remained flat absent treatment. Pre-treatment event studies frequently reject this flat-gap requirement. Existing responses include parametric trend controls and bounds on the treatment effect under assumptions about the magnitude of the violation. This paper shows that point identification of cohort-specific and aggregate treatment effects in staggered designs remains achievable under strictly weaker assumptions. I replace the flat-gap requirement with a hierarchy of higher-order conditions, Parallel[p], embed this framework in the group-time average treatment effect structure of Callaway and Sant'Anna (2021), and prove an aggregation theorem for the case where different cohorts are identified under different feasible polynomial orders, a challenge unique to staggered designs that has not been previously addressed. A sequential order-selection procedure guides applied practice. Monte Carlo evidence confirms that post-selection bootstrap coverage remains near-nominal and that inference is robust to realistic serial correlation. Applied to Medicaid expansion data, the method yields point estimates resting on an assumption the pre-treatment data do not reject, in contrast to the flat-gap requirement which those same data decisively reject.

Interference poses a significant challenge to satellite-based positioning systems, making it essential to accurately quantify the effects of specific interference types on receiver performance and the resulting reliability of position computation. In current practice, interference effects are often quantified using receiver-independent metrics, with receiver-specific front-end characteristics either idealized or only implicitly considered. In this paper, we address this limitation by explicitly incorporating receiver-specific front-end characteristics into the computation of interference effects and validating the resulting receiver-dependent analysis experimentally. Therefore, we record a real-world open-field dataset comprising 210 distinct interference scenarios and compute the receiver-dependent spectral separation coefficient (SSC) and interference impact for a specific receiver module. Furthermore, we verify the computation using a controlled dataset generated with a radio frequency constellation simulator (RFCS), employing the same receiver module and replaying similar interferences classes. The comparison of results obtained in both environments demonstrates the robustness of the interference impact computation.

We propose diarization-conditioned spoken language models (SLMs), a strategy for extending SLMs to far-field multi-talker audio. Rather than adapting the decoder via Serialized Output Training, which risks catastrophic forgetting, we condition the acoustic encoder on diarization masks to extract target-speaker representations, keeping the decoder frozen. We instantiate this as Dixtral, integrating a Diarization Conditioned Whisper (DiCoW) encoder into the Voxtral SLM. On AMI, NOTSOFAR-1, LibriSpeechMix, and Mixer6, Dixtral outperforms Gemini 3.0 Flash, VibeVoice, and Voxtral Mini Transcribe V2 on speaker-attributed transcription by 29.0%, 19.8%, and 16.0% absolute cpWER respectively. On a novel long-form multi-speaker QA benchmark, zero-shot Dixtral matches Gemini on far-field content understanding, and when fine-tuned surpasses both Gemini and Voxtral operating on close-talk across all tasks.

Neural audio codecs are central to modern LLM-based Text-to-Speech (TTS) and multimodal systems. As low-bitrate semantic codecs gain prominence, the Token-to-Waveform (Token2Wav) decoder becomes a bottleneck determining both perceptual quality and system efficiency. Conventional multi-step flow-matching decoders offer superior quality but suffer from high inference latency due to iterative sampling, creating a severe quality-speed trade-off. In this paper, we propose a novel Token2Wav architecture that overcomes this limitation by applying MeanFlow in a highly compressed latent space. By modeling the average velocity rather than the instantaneous velocity field, MeanFlow enables true one-step generation. Operating in the latent domain mitigates the memory and stability issues of waveform-level flows, yielding up to a 17$\times$ improvement in Real-Time Factor (RTF) compared to multi-step baselines with negligible quality degradation. Furthermore, we introduce refinement strategies that mitigate latent mismatch, including decoder-only fine-tuning with the MeanFlow generator frozen and end-to-end joint fine-tuning, improving fidelity without increasing inference-time cost. Code and demo are publicly available.

Picture descriptions provide valuable insights into several clinical constructs related to cognitive-linguistic abilities. However, operationalizing these constructs into quantitative measures remains challenging, limiting interpretability and clinical utility. We introduced seven constructs tailored to the Cookie Theft picture description task and prompted large language models (LLMs) to evaluate them, generating severity scores and example-based explanations. Among the examined LLMs, Claude 3.5 Sonnet performed the best, producing severity scores that significantly distinguish cognitively impaired individuals from healthy controls. The model achieves a high accuracy of 85% on the ADReSS dataset. Expert evaluation of Claude's scores and explanations yields a 3.99/5 average agreement. The findings demonstrate the potential of LLMs to operationalize clinical constructs and generate interpretable evaluations, offering a promising approach for accessible cognitive screening tools.

We investigate the optimum energy per bit in hollow-core-fibre transmission systems. We show that a 1000 km C-band link can achieve a 41.5% reduction in total power consumption when operating at the minimum energy-per-bit launch power with only 2.2% throughput penalty.

Unified receivers (URs) have emerged as a promising architecture for simultaneous wireless information and power transfer (SWIPT), since a common rectifying front-end enables information decoding (ID) and energy harvesting (EH) from the same rectified output. However, rectification is nonlinear due to the diode, while the capacitor introduces memory across symbols, making constellation design over the channel challenging. In this paper, we study constellation design for nonlinear UR-SWIPT channels in both memoryless and memory regimes. First, we propose a tractable unified rectification model that captures both (i) the nonlinear steady-state mapping and (ii) the asymmetric capacitor charging/discharging dynamics under transient operation. To isolate the impact of rectification with memory on ID, we study the information-based design. In this setting, we develop a state-adaptive policy with an algorithmic constellation design that accounts for the rectifier state and shapes the constellation in the observation domain. By approximating the rectifier state distribution, we derive a closed-form average symbol error rate (SER) expression and characterize the rate-reliability (R-R) tradeoff. We then seek constellations that minimize the SER under average transmit power and EH constraints. We address the resulting energy-constrained setting in the memoryless regime using an autoencoder-based framework that embeds the nonlinear rectification model as a differentiable channel block. Numerical results validate the proposed models, demonstrate the impact of memory on the R-R tradeoff, and show how learned constellations adapt to EH requirements in the rate-energy tradeoff.

Over-the-air computation (AirComp) has emerged as a promising approach for massive data aggregation, which is yet challenged by the channel variations, task distributions, and inherent energy limitation of the computation nodes. In this paper, we propose an unmanned aerial vehicle (UAV)-assisted Aircomp system to serve multi-cluster computation tasks over time, where the UAV mobility-facilitated spatial and time diversity is exploited for efficient and accurate data computation. Specifically, we aim for the minimization of AirComp aggregation error and the energy consumption by jointly optimizing the transceiver beamforming, normalizing factors, sensor scheduling, and UAV trajectory. To solve the formulated problem, we decompose it into two layers where the inner layer addresses the optimization-based AirComp transceiver design, and the outer layer focuses on the deep reinforcement learning (DRL)-based scheduling and trajectory design. In particular, a pointer network actor-critic learning is developed to tackle the binary scheduling problem, and a soft actor-critic DRL algorithm is employed to determine the UAV trajectory. Simulation results validate the convergence of the proposed hierarchical learning framework and demonstrate its significant performance gains in terms of AirComp aggregation error and energy consumption as compared with baseline schemes.

Learned communication systems may evaluate stochastic channel surrogates millions of times inside differentiable training loops, making diffusion-style reverse sampling expensive. This paper proposes condition-wise Sinkhorn drifting, a one-shot channel surrogate that preserves the transmitted symbol and transports only the conditional output laws \(p(y\mid x)\). We formulate a conditional Sinkhorn objective over repeated outputs at the same transmitted symbol and train the generator with finite-sample barycentric velocities followed by detached particle regression. Experiments on additive white Gaussian noise (AWGN), Rayleigh fading, solid-state power amplifier (SSPA) nonlinearity, and a compact tapped-delay-line (TDL) channel compare direct drifting, joint Sinkhorn drifting, condition-wise Sinkhorn drifting, conditional denoising diffusion probabilistic modeling (DDPM), denoising diffusion implicit modeling (DDIM), and Wasserstein generative adversarial network (WGAN) references. Within the evaluated one-shot drifting-family variants, condition-wise Sinkhorn is strongest under conditional diagnostics and symbolic-coding checks, while diffusion remains strongest on the hardest downstream symbol-error-rate (SER) curves. The resulting operating point is a condition-preserving one-shot simulator for settings where repeated channel calls make diffusion-style sampling too costly.

Improvements in the dynamic range and sensitivity of digital MEMS microphones are essential in applications like advanced noise canceling and voice recognition. A cost effective solution to achieve these goals is the companding ADC architecture. Companding ADCs split the dynamic range in several segments with different quantization noise levels, relaxing power constraints. A common problem of companding microphones are audible artifacts generated when the input signal crosses the boundaries between different amplitude segments. We show in this paper a companding ADC architecture that mitigates the boundary artifacts by leveraging the instantaneous and high-resolution time-domain representation of the input signal in a VCO-based ADC. The use of a multi-rate frequency-to-digital converter allows to decouple quantization noise from the VCO frequency, keeping standard audio sampling rates. Co-optimization of the driver and oscillator circuits enables our VCO-ADC to reach \textgreater 112dBc of peak SFDR without a feedback DAC, keeping a Giga-Ohm input impedance compatible with a capacitive MEMS. We show measurements of a 0.13 $μ$m ASIC implementing a complete readout circuit for a digital MEMS microphone. This includes two analog channels and the digital signal processing and calibration blocks required to deliver a standard single-bit PDM output. This ADC reaches a dynamic range of 114.3dB with a power budget under 400 uW, a Schreier FoM_{SNDR} of 171.0 dB and a FoM_{DR} of 191.3 dB.

Objective video quality metrics commonly assume uniform spatial attention, an assumption that conflicts with the selective nature of human visual perception, particularly in sports videos. Here, allocating more bits for salient regions through semantic encoding can lead to significant bitrate savings. We present a Perceptually-Weighted Video Quality Metric (PW-VQM), a full-reference metric that accounts for the unequal perceptual importance of spatial regions and therefore targets quality evaluation for asymmetrically encoded content. SSIM maps computed in a multiscale wavelet domain are weighted by differentiating between foreground and background regions. Perceptually salient foreground regions are identified by combining open-vocabulary object detection with optical flow analysis, and are assigned higher weight during quality aggregation. Evaluated on sports video content, PW-VQM achieves a Spearman Rank Order Correlation Coefficient of 0.9511, outperforming established metrics including SSIM, VMAF, FUNQUE, and LPIPS. An ablation study confirms the individual contributions of the components of the perceptual weighting.

Flow matching (FM) enables high-fidelity generation, while self-supervised learning (SSL) speech models provide hierarchical representations spanning acoustic and phonetic levels. However, existing FM-based speech enhancement (SE) methods operate primarily in the spectral domain, treating SSL features only as external conditions rather than modeling directly in the SSL latent space. To fully exploit the structural richness of SSL representations, we propose PhASE-Flow, an FM-based SE framework that operates entirely in the SSL space. It models the conditional distribution of clean acoustic representations given phonetic ones, reconstructing the waveform via a neural vocoder. Experiments show that PhASE-Flow outperforms state-of-the-art baselines in perceptual quality and intelligibility. Notably, it achieves competitive performance with only four sampling steps, enabling highly efficient inference. Audio demos are available at https://anonymous.4open.science/w/phase-flow_demo-E6E1/.

An FFT-based direction-of-arrival (DOA) and range-estimation framework for a monostatic uniform linear array (ULA) operating at 77 GHz is presented. A narrowband sinusoidal waveform is used to derive the spatial phase model, determine an aliasing-free inter-element spacing, and select the aperture required to obtain a boresight angular resolution of 2 degree. The resulting design uses an element spacing of 0.97 mm and 58 antenna elements, corresponding to an aperture length of 56.42 mm. Numerical results show accurate angular estimation for a single target at 30 degree and for multiple simultaneous targets. The analysis is further extended to two-dimensional localization by replacing the narrowband waveform with a 1 GHz sinc-modulated signal, which provides an approximate range resolution of 0.15 m. Additional simulations quantify the effects of additive complex Gaussian noise, increased antenna spacing, and target decorrelation on the DOA response.

Downlink channel state information (CSI) feedback plays a key role in frequency division duplex (FDD) massive multiple-input multiple-output (mMIMO) systems. The growth of antennas in ultra-massive MIMO increases the difficulty and overhead of CSI feedback, which poses significant challenges for conventional downlink CSI feedback mechanisms. To address the limitations of existing CSI feedback approaches, this paper proposes a novel curvelet learning based framework termed SwinCANet, comprising a frequency-domain information processing module and a denoising module. The frequency-domain information processing module employs curvelet transform to decompose CSI into low-frequency and high-frequency components. Subsequently, Swin Transformer and channel-wise attention block are utilized for extracting the low-frequency and high-frequency representations, respectively, thereby enhancing reconstruction quality. Notably, an additional Swin Transformer facilitates the fusion of multi-scale frequency components, enhancing capabilities across different angular resolutions and spatial directions. Furthermore, we develop a variant (De-SwinCANet), which employs a Sigmoid threshold function to effectively suppress noise coefficients, thereby mitigating various channel impairments and nonlinear distortions. Numerical simulation results demonstrate that the proposed methodology achieves superior performance compared to existing benchmarks while maintaining robust performance under challenging propagation conditions.

Integrated Sensing and Communications (ISAC) is a defining feature of 6G, extending cellular networks with radar-like sensing at limited additional overhead. In bistatic deployments, sensing requires coordinating the transmitter (TX) and receiver (RX) arrays to scan the Cartesian product of angle of departure and arrival, resulting in a four-dimensional sampling problem in the angular domain. This work establishes a complete angular sampling framework for bistatic ISAC, extending the DFT-based optimal-sampling methodology to the full azimuth and elevation domains of both arrays. We show that the bistatic geometry couples the TX and RX elevation angles, and represent this coupling through the ortho-baseline coarray, a virtual array that captures the joint elevation aperture of the array pair. From the coarray we derive a minimal sampling and interpolation scheme, near-lossless and realizable with any beamforming architecture. Monte Carlo simulations confirm the proposed minimal acquisition essentially equalizes the detection accuracy of dense oversampled imaging while acquiring 3 to 5 times fewer TX-RX direction pairs. This allows having bistatic operations with drastically reduced overhead on the radio resource usage of ISAC systems.

We propose a novel approach to the synchronization paradigm in distributed ISAC (DISAC) systems in doubly-dispersive (DD) channel environments via a joint synchronization and radar parameter estimation framework. The proposed method exploits the structure of the system model, which can be linearized in order to apply a bivariate Gaussian belief propagation (GaBP) algorithm that jointly estimates the time offset (TO) and carrier frequency offset (CFO) of each base station (BS), as well as the delay and Doppler parameters of the DD channel in conventional orthogonal frequency division multiplexing (OFDM) systems. Simulation results demonstrate the effectiveness of the proposed algorithm, showing that the radar parameter estimates (i.e., range and velocity) and synchronization parameter estimates (i.e., TO and CFO) approach the Cramér Rao lower bound (CRLB) even at low-to-moderate signal-to-noise ratio (SNR) regimes.

Synthetic data has the potential to be a valuable resource for training machine learning models, particularly Automatic Speech Recognition (ASR) Systems; however, its effectiveness requires systematic evaluation. In this study, we investigate the impact of incorporating synthetic speech data alongside real-world recordings for three Indic languages: Hindi, Kannada, and Telugu. We analyze the performance gains achieved by augmenting synthetic data with real data and independently examine how ASR performance varies with the sources of scripts used to generate synthetic speech. In addition, we evaluate the effect of synthetic speech generated using different speech synthesis models. Finally, we study the impact of voice cloning in synthetic speech generation on ASR performance, including how performance varies with the number of distinct cloned voices used during data generation.

In this paper, we investigate an active-RIS (ARIS)-aided integrated sensing and communication (ISAC) system with Rydberg Atomic REceiver (RARE). Leveraging the magnitude-only and real-domain observation structure of RARE, we first derive a unified ISAC model, along with a closed-form Cramer-Rao bound (CRB) for direction-of-arrival (DoA) estimation. Based on this formulation, we propose a joint design of the {base station (BS)} beamforming and ARIS reflection coefficients to minimize the CRB under RARE-specific signal-to-interference-noise-ratio (SINR) and ARIS power constraints. To tackle the resulting highly non-convex problem, we develop an alternating optimization (AO) framework that combines semidefinite relaxation (SDR) for beamforming and a majorization-minimization (MM)-based approach for ARIS design. Numerical results demonstrate that the proposed RARE-aware framework significantly outperforms conventional RF-based designs and achieves performance close to the radar-only benchmark, highlighting the potential of RARE for quantum-enhanced ISAC with ARIS.

Introduction: To develop and evaluate Fine-UNETR, a Vision Transformer-based architecture for automated segmentation of PSMA-avid lesions on whole-body PET/CT, and to assess clinical utility of AI-derived tumor burden biomarkers for overall survival stratification in radioligand therapy. Methods: In this retrospective study, 373 PSMA PET/CT scans (mean age, 71+-8 years) from patients with prostate cancer were analyzed. Fine-UNETR, a modified UNETR with 8x8x8 voxel patch embedding and axial sliding window training, was trained on 299 scans and validated on 74 scans. Overall survival stratification was assessed in an independent cohort of 67 pre-radioligand therapy patients using Kaplan-Meier analysis and log-rank testing. External validation was performed on 192 cases from the AutoPET IV PSMA PET/CT dataset. Results: Fine-UNETR achieved a Dice similarity coefficient (DSC) of 66.63%, sensitivity of 70.27%, precision of 67.77%, and a lesion detection rate of 79.53% (96.05% for lesions with SUVmax >= 5). On the external validation dataset, the model achieved a DSC of 44.11% and a lesion detection rate of 87.18%, indicating that lesion detection performance was preserved despite reduced voxel-level overlap. AI-derived biomarkers showed excellent agreement with ground truth (total tumor volume: r=0.984; total lesion uptake: r=0.989; lesion count: r=0.960). In the clinical cohort, total tumor volume (p=0.0019), SUVmax (p=0.014), and SUVmean (p=0.016) significantly stratified overall survival. Conclusion: Fine-UNETR enables accurate automated whole-body PSMA lesion segmentation and tumor burden quantification. Performance on an external dataset demonstrates robustness despite evidence of domain shift. AI-derived biomarkers significantly stratified overall survival in a pre-radioligand therapy cohort, supporting the clinical utility of automated PSMA PET/CT quantification for prognostication.

A compact dynamic omnidirectional array is proposed for antenna-level physical-layer security through directional modulation. Unlike conventional directional-modulation transmitters based on phased-array beam synthesis or multiple RF chains, the proposed architecture uses a single RF input and a switching-controlled four-element printed meander-line monopole array operating at 5.05 GHz. The state-dependent excitation introduces controllable magnitude and phase perturbations in the radiated field, producing angle-dependent constellation distortion and bit error rate behavior. Reliable information recovery is confined to a narrow broadside region in the E-plane, whereas the H-plane remains quasi-static and omnidirectional, providing a full 360-degree information-recoverable region. The antenna is implemented on a single-layer Rogers RO4350B substrate with a compact footprint of 0.57 x 1.11 lambda_0^2. A four-path switching network based on commercial RF components is used for experimental validation. Communication measurements using 16-QAM at 5.05 GHz demonstrate BER-defined E-plane information beamwidths of 30 to 36 degrees for calibrated switching modes under a BER <= 10^-3 criterion, while no bit errors are observed in the measured H-plane and the SNR remains above approximately 33 dB. Feed-phase offsets are also used to steer the BER-defined information-recoverable sector, demonstrating information-beam steering with the same antenna-level switching mechanism. These results show that compact antenna-level directional modulation can provide angularly selective information recovery in one principal plane while preserving omnidirectional coverage in the orthogonal plane.

Affine frequency division multiplexing (AFDM) is an emerging chirp-based multicarrier waveform with strong diversity in doubly selective channels, but practical systems suffer from transmitter and receiver IQ imbalance, causing image interference and performance degradation. This paper proposes a two-stage IQ imbalance estimation and compensation method for AFDM systems. First, a preamble-assisted iterative algorithm estimates the time-invariant IQ imbalance parameters by exploiting their slowly time-varying nature. Then, a joint channel estimation and data detection scheme combines basis expansion model (BEM)-based channel estimation with an improved LMMSE detector for interference suppression. Simulations show rapid convergence and near-ideal BER performance.

Electrochemical Impedance Spectroscopy (EIS) is a widely used, non-destructive technique for characterizing electrochemical systems, and its analysis typically relies on fitting the measured spectra to an Equivalent Circuit Model (ECM). Selecting an appropriate ECM, however, remains a major bottleneck: knowledge-based selection requires expert judgment and is difficult to reproduce, while existing automated approaches either choose from a fixed set of candidate circuits or, in the case of Gene Expression Programming, require repeated equivalent-circuit fitting and a predetermined circuit scale. Here, we propose a machine learning method that estimates an ECM directly from an impedance spectrum by representing the circuit as a serialized string of symbols and generating this string with a Long Short-Term Memory (LSTM) network coupled to a convolutional feature extractor. Because the LSTM inherently handles variable-length sequences, the method produces the circuit topology directly, without any fitting during estimation nor prior assumption for the number of elements. A fourth-root transformation of the impedance is introduced to emphasize the mid-frequency features essential for distinguishing circuits, and an adaptive beam search yields multiple ranked candidates. Evaluated on 100,000 synthetic datasets generated from 119 circuit topologies with 1% added noise on impedances, the method identified the correct topology as the most probable ECM in 77.8% of cases and among the top five candidates in 98.8% of cases, with an average estimation time of 17.8 milliseconds per dataset - several orders of magnitude faster than reported fitting-based approaches. These results indicate that direct topology generation with a neural network is a promising route toward fully automated, expert-independent ECM estimation.

In this study, we focus on cough-based tuberculosis screening (CBTS) and hypothesize that fusing speech/audio foundation representations with spectral descriptors will yield stronger screening performance. We expect this fusion to reveal complementary strengths: spectral features preserve fine-grained short-time acoustic detail in cough signals, while foundation embeddings capture higher-level temporal and event-level patterns learned from large-scale pretraining. To this end, we propose COBALT, a novel fusion framework based on codebook-aligned hyperbolic prototypes and bandit-style reliability weighting to integrate heterogeneous representations effectively. Using the CODA TB DREAM Challenge benchmark, COBALT consistently outperforms individual representations and a concatenation baseline, achieving the best overall performance when fusing MFCC with PaSST thereby establishing a new state-of-the-art on the benchmark.

Abscisic acid (ABA) is a central plant hormone for coordinating responses to drought, salinity, cold stress, pathogen attack, wounding, and developmental aging. This paper reviews the biological stimuli that increase ABA biosynthesis, the main production sites and pathways, and the long-distance movement of ABA through plant vascular tissues. It then discusses experimental quantification approaches, including gas-liquid chromatography with electron-capture detection and high-performance liquid chromatography with ultraviolet detection. Finally, the paper presents a molecular-communication-inspired model of ABA transport in which root-side ABA release is represented as a transmitter, the xylem pathway as a bounded channel, and soybean tissue as a receiver. MATLAB Brownian-motion simulations are used to evaluate the effects of released molecule quantity and receiver radius on the detected ABA signal. The results show that higher release quantities produce smoother and stronger reception trends, while larger receivers increase molecule-capture probability.