Show newer

Do Segmentation Models Understand Vascular Structure? A Blob-Based XAI Framework arxiv.org/abs/2504.11469 .IV .AI .CV

Do Segmentation Models Understand Vascular Structure? A Blob-Based XAI Framework

Deep learning models have achieved impressive performance in medical image segmentation, yet their black-box nature limits clinical adoption. In vascular applications, trustworthy segmentation should rely on both local image cues and global anatomical structures, such as vessel connectivity or branching. However, the extent to which models leverage such global context remains unclear. We present a novel explainability pipeline for 3D vessel segmentation, combining gradient-based attribution with graph-guided point selection and a blob-based analysis of Saliency maps. Using vascular graphs extracted from ground truth, we define anatomically meaningful points of interest (POIs) and assess the contribution of input voxels via Saliency maps. These are analyzed at both global and local scales using a custom blob detector. Applied to IRCAD and Bullitt datasets, our analysis shows that model decisions are dominated by highly localized attribution blobs centered near POIs. Attribution features show little correlation with vessel-level properties such as thickness, tubularity, or connectivity -- suggesting limited use of global anatomical reasoning. Our results underline the importance of structured explainability tools and highlight the current limitations of segmentation models in capturing global vascular context.

arXiv.org

Local Temporal Feature Enhanced Transformer with ROI-rank Based Masking for Diagnosis of ADHD arxiv.org/abs/2504.11474 .IV .AI .CV

Local Temporal Feature Enhanced Transformer with ROI-rank Based Masking for Diagnosis of ADHD

In modern society, Attention-Deficit/Hyperactivity Disorder (ADHD) is one of the common mental diseases discovered not only in children but also in adults. In this context, we propose a ADHD diagnosis transformer model that can effectively simultaneously find important brain spatiotemporal biomarkers from resting-state functional magnetic resonance (rs-fMRI). This model not only learns spatiotemporal individual features but also learns the correlation with full attention structures specialized in ADHD diagnosis. In particular, it focuses on learning local blood oxygenation level dependent (BOLD) signals and distinguishing important regions of interest (ROI) in the brain. Specifically, the three proposed methods for ADHD diagnosis transformer are as follows. First, we design a CNN-based embedding block to obtain more expressive embedding features in brain region attention. It is reconstructed based on the previously CNN-based ADHD diagnosis models for the transformer. Next, for individual spatiotemporal feature attention, we change the attention method to local temporal attention and ROI-rank based masking. For the temporal features of fMRI, the local temporal attention enables to learn local BOLD signal features with only simple window masking. For the spatial feature of fMRI, ROI-rank based masking can distinguish ROIs with high correlation in ROI relationships based on attention scores, thereby providing a more specific biomarker for ADHD diagnosis. The experiment was conducted with various types of transformer models. To evaluate these models, we collected the data from 939 individuals from all sites provided by the ADHD-200 competition. Through this, the spatiotemporal enhanced transformer for ADHD diagnosis outperforms the performance of other different types of transformer variants. (77.78ACC 76.60SPE 79.22SEN 79.30AUC)

arXiv.org

Accelerated Recovery with RIS: Designing Wireless Resilience in Mission-Critical Environments arxiv.org/abs/2504.11589 .SP .SY .SY

Accelerated Recovery with RIS: Designing Wireless Resilience in Mission-Critical Environments

As 6G and beyond redefine connectivity, wireless networks become the foundation of critical operations, making resilience more essential than ever. With this shift, wireless systems cannot only take on vital services previously handled by wired infrastructures but also enable novel innovative applications that would not be possible with wired systems. As a result, there is a pressing demand for strategies that can adapt to dynamic channel conditions, interference, and unforeseen disruptions, ensuring seamless and reliable performance in an increasingly complex environment. Despite considerable research, existing resilience assessments lack comprehensive key performance indicators (KPIs), especially those quantifying its adaptability, which are vital for identifying a system's capacity to rapidly adapt and reallocate resources. In this work, we bridge this gap by proposing a novel framework that explicitly quantifies the adaption performance by augmenting the gradient of the system's rate function. To further enhance the network resilience, we integrate Reconfigurable Intelligent Surfaces (RISs) into our framework due to their capability to dynamically reshape the propagation environment while providing alternative channel paths. Numerical results show that gradient augmentation enhances resilience by improving adaptability under adverse conditions while proactively preparing for future disruptions.

arXiv.org

Provably Safe Control for Constrained Nonlinear Systems with Bounded Input arxiv.org/abs/2504.11592 .SY .DS .OC .SY

Provably Safe Control for Constrained Nonlinear Systems with Bounded Input

In real-world control applications, actuator constraints and output constraints (specifically in tracking problems) are inherent and critical to ensuring safe and reliable operation. However, generally, control strategies often neglect these physical limitations, leading to potential instability, degraded performance, or even system failure when deployed on real-world systems. This paper addresses the control design problem for a class of nonlinear systems under both actuator saturation and output constraints. First, a smooth asymmetric saturation model (a more generic representative of practical scenarios) is proposed to model actuator saturation, which ensures that the control inputs always remain confined within a predefined set to ensure safety. Based on the proposed model, we develop a nonlinear control framework that guarantees output tracking while ensuring that system output remains confined to the predefined set. Later, we integrate this design with the constrained output tracking control problem, wherein we show that the system output tracks its desired trajectory by simultaneously satisfying input and output constraints. The global stabilization of the tracking error is achieved in the presence of input constraints, while semi-global stabilization is achieved in the presence of both input and output constraints. Additionally, we rigorously establish the boundedness of all closed-loop signals under the proposed design. Simulation results demonstrate the effectiveness of the proposed methods in handling asymmetric constraints while achieving desirable tracking performance.

arXiv.org

Integrating electrocardiogram and fundus images for early detection of cardiovascular diseases arxiv.org/abs/2504.10493 .IV .CV

Integrating electrocardiogram and fundus images for early detection of cardiovascular diseases

Cardiovascular diseases (CVD) are a predominant health concern globally, emphasizing the need for advanced diagnostic techniques. In our research, we present an avant-garde methodology that synergistically integrates ECG readings and retinal fundus images to facilitate the early disease tagging as well as triaging of the CVDs in the order of disease priority. Recognizing the intricate vascular network of the retina as a reflection of the cardiovascular system, alongwith the dynamic cardiac insights from ECG, we sought to provide a holistic diagnostic perspective. Initially, a Fast Fourier Transform (FFT) was applied to both the ECG and fundus images, transforming the data into the frequency domain. Subsequently, the Earth Mover's Distance (EMD) was computed for the frequency-domain features of both modalities. These EMD values were then concatenated, forming a comprehensive feature set that was fed into a Neural Network classifier. This approach, leveraging the FFT's spectral insights and EMD's capability to capture nuanced data differences, offers a robust representation for CVD classification. Preliminary tests yielded a commendable accuracy of 84 percent, underscoring the potential of this combined diagnostic strategy. As we continue our research, we anticipate refining and validating the model further to enhance its clinical applicability in resource limited healthcare ecosystems prevalent across the Indian sub-continent and also the world at large.

arXiv.org

Remote Sensing Based Crop Health Classification Using NDVI and Fully Connected Neural Networks arxiv.org/abs/2504.10522 .IV .CV

Remote Sensing Based Crop Health Classification Using NDVI and Fully Connected Neural Networks

Accurate crop health monitoring is not only essential for improving agricultural efficiency but also for ensuring sustainable food production in the face of environmental challenges. Traditional approaches often rely on visual inspection or simple NDVI measurements, which, though useful, fall short in detecting nuanced variations in crop stress and disease conditions. In this research, we propose a more sophisticated method that leverages NDVI data combined with a Fully Connected Neural Network (FCNN) to classify crop health with greater precision. The FCNN, trained using satellite imagery from various agricultural regions, is capable of identifying subtle distinctions between healthy crops, rust-affected plants, and other stressed conditions. Our approach not only achieved a remarkable classification accuracy of 97.80% but it also significantly outperformed conventional models in terms of precision, recall, and F1-scores. The ability to map the relationship between NDVI values and crop health using deep learning presents new opportunities for real-time, large-scale monitoring of agricultural fields, reducing manual efforts, and offering a scalable solution to address global food security.

arXiv.org

Imaging Transformer for MRI Denoising: a Scalable Model Architecture that enables SNR << 1 Imaging arxiv.org/abs/2504.10534 .med-ph .IV .SP

Imaging Transformer for MRI Denoising: a Scalable Model Architecture that enables SNR << 1 Imaging

Purpose: To propose a flexible and scalable imaging transformer (IT) architecture with three attention modules for multi-dimensional imaging data and apply it to MRI denoising with very low input SNR. Methods: Three independent attention modules were developed: spatial local, spatial global, and frame attentions. They capture long-range signal correlation and bring back the locality of information in images. An attention-cell-block design processes 5D tensors ([B, C, F, H, W]) for 2D, 2D+T, and 3D image data. A High Resolution (HRNet) backbone was built to hold IT blocks. Training dataset consists of 206,677 cine series and test datasets had 7,267 series. Ten input SNR levels from 0.05 to 8.0 were tested. IT models were compared to seven convolutional and transformer baselines. To test scalability, four IT models 27m to 218m parameters were trained. Two senior cardiologists reviewed IT model outputs from which the EF was measured and compared against the ground-truth. Results: IT models significantly outperformed other models over the tested SNR levels. The performance gap was most prominent at low SNR levels. The IT-218m model had the highest SSIM and PSNR, restoring good image quality and anatomical details even at SNR 0.2. Two experts agreed at this SNR or above, the IT model output gave the same clinical interpretation as the ground-truth. The model produced images that had accurate EF measurements compared to ground-truth values. Conclusions: Imaging transformer model offers strong performance, scalability, and versatility for MR denoising. It recovers image quality suitable for confident clinical reading and accurate EF measurement, even at very low input SNR of 0.2.

arXiv.org

Secure Estimation of Battery Voltage Under Sensor Attacks: A Self-Learning Koopman Approach arxiv.org/abs/2504.10639 .SY .SY

Secure Estimation of Battery Voltage Under Sensor Attacks: A Self-Learning Koopman Approach

Cloud-based battery management system (BMS) requires accurate terminal voltage measurement data to ensure optimal and safe charging of Lithium-ion batteries. Unfortunately, an adversary can corrupt the battery terminal voltage data as it passes from the local-BMS to the cloud-BMS through the communication network, with the objective of under- or over-charging the battery. To ensure accurate terminal voltage data under such malicious sensor attacks, this paper investigates a Koopman-based secure terminal voltage estimation scheme using a two-stage error-compensated self-learning feedback. During the first stage of error correction, the potential Koopman prediction error is estimated to compensate for the error accumulation due to the linear approximation of Koopman operator. The second stage of error compensation aims to recover the error amassing from the higher-order dynamics of the Lithium-ion batteries missed by the self-learning strategy. Specifically, we have proposed two different methods for this second stage error compensation. First, an interpretable empirical correction strategy has been obtained using the open circuit voltage to state-of-charge mapping for the battery. Second, a Gaussian process regression-based data-driven method has been explored. Finally, we demonstrate the efficacy of the proposed secure estimator using both empirical and data-driven corrections.

arXiv.org

Correcting Domain Shifts in Electric Motor Vibration Data for Unseen Operating Conditions arxiv.org/abs/2504.10661 .SP

Correcting Domain Shifts in Electric Motor Vibration Data for Unseen Operating Conditions

This paper addresses the problem of domain shifts in electric motor vibration data created by new operating conditions in testing scenarios, focusing on bearing fault detection and diagnosis (FDD). The proposed method combines the Harmonic Feature Space (HFS) with regression to correct for frequency and energy differentials in steady-state data, enabling accurate FDD on unseen operating conditions within the range of the training conditions. The HFS aligns harmonics across different operating frequencies, while regression compensates for energy variations, preserving the relative magnitude of vibrations critical for fault detection. The proposed approach is evaluated on a detection problem using experimental data from a Belt-Starter Generator (BSG) electric motor, with test conditions having a minimum 1000 RPM and 5 Nm difference from training conditions. Results demonstrate that the method outperforms traditional analysis techniques, achieving high classification accuracy at a 94% detection rate and effectively reducing domain shifts. The approach is computationally efficient, requires only healthy data for training, and is well-suited for real-world applications where the exact application operating conditions cannot be predetermined.

arXiv.org

Spectrum Sharing in STAR-RIS-assisted UAV with NOMA for Cognitive Radio Networks arxiv.org/abs/2504.10691 .SY .SY

Spectrum Sharing in STAR-RIS-assisted UAV with NOMA for Cognitive Radio Networks

As an emerging technology, the simultaneous transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) can improve the spectrum efficiency (SE) of primary users (PUs) and secondary users (SUs) in cognitive radio (CR) networks by mitigating the interference of the incident signals. The STAR-RIS-assisted unmanned aerial vehicle (UAV) can fully cover the dynamic environment through high mobility and fast deployment. According to the dynamic air-to-ground channels, the STAR-RIS-assisted UAV may face a challenge configuring their elements' coefficients (i.e., reflecting and transmitting the amplitude and phases). Hence, to meet the requirements of dynamic channel determination with the SE approach, this paper proposes the sum rate maximization of both PUs and SUs through non-orthogonal multiple access in CR network to jointly optimize the trajectory and transmission-reflection beamforming design of the STAR-RIS-assisted UAV, and power allocation. Since the non-convex joint optimization problem includes coupled optimization variables, we develop an alternative optimization algorithm. Simulation results study the impact of: 1) the significant parameters, 2) the performance of different intelligence surface modes and STAR-RIS operating protocols, 3) the joint trajectory and beamforming design with fixed and mobile users, and 4) STAR-RIS capabilities such as mitigating the interference, and how variations in the roles of elements dynamically.

arXiv.org
Show older
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.