IEEE Access

The increasing availability of portable ultrasound systems motivates exploration of novel approaches to respiratory signal assessment. In this in-vitro study, we investigate whether pulsed-wave (PW) Doppler ultrasound can capture structured spectral patterns from replayed lung sound recordings. Digitized respiratory sounds were replayed through a tissue-mimicking ultrasound phantom, generating 1,478 PW Doppler spectral images from recordings associated with healthy subjects and several externally labeled disease categories. Exploratory classification experiments using a ResNet-18 architecture demonstrated that these Doppler representations contain learnable differences under controlled conditions. These findings motivate further investigation into PW Doppler as a potential representation of respiratory acoustics.

Synchrony is a key mechanism that builds up the foundations of human interactions. Quantifying the level of physiological synchronization that occurs during dyadic exchanges is essential to fully comprehend social phenomena. We present a new index to characterize the coupling of complex physiological dynamics: the optimized Multichannel Complexity Index (opMCI). We validated this approach using synthetic time series of two coupled Henon Maps, with four different coupling levels in unidirectional and bidirectional manners. We demonstrated that the opMCI method allows to effectively discern between all coupling levels. Then, we applied the opMCI metric on heart rate variability data collected from 37 parent-infant dyads, during shared reading and playing activities, in the framework of the Shared Emotional Reading (SHER) project, with the aim of assessing the effects of early intervention in preterm babies. Two groups presented preterm infants: an intervention group, who participated in a two-month shared reading program, and a control group, who practiced shared play activities. A full-term group provided additional control data. The opMCI values were significantly higher for the intervention dyads with respect to the other groups during the shared reading task, showing that an early reading intervention program could increase parent-infant synchrony in preterm babies.

In "Experiment-free exoskeleton assistance via learning in simulation", Luo et al. [1] present an ambitious framework for developing exoskeleton controllers through reinforcement learning exclusively in computer simulation. The authors report that a control policy trained on a small dataset from one subject was directly transferred to physical hardware, reducing human metabolic cost during walking, running, and stair climbing by more than any prior device. If confirmed, this would represent a major breakthrough for the field of wearable robotics and their clinical applications. However, a close examination of the published materials casts doubt on these claims. The reported experimental results violate physiological limits on the relationship between mechanical power and muscle energy use during gait2,3,4. The algorithmic claims are surprising and cannot be verified; in contrast with established replicability standards in machine learning5,6, executable code has not been made available. We conclude that the goals of this study have not yet been verifiably achieved and make recommendations for avoiding publication errors of this type in the future.

Background and ObjectiveVariational Autoencoders (VAEs) offer a powerful framework for unsupervised anomaly detection and data clustering, often surpassing traditional methods. A core strength of VAEs lies in their ability to model data distributions probabilistically, enabling robust identification of anomalies and clusters through reconstruction likelihood -- a stochastic metric providing a principled alternative to deterministic error scores. MethodsWe investigated how different VAE architectures, combining reconstruction likelihood with a learnable or data-driven prior, performed in a clustering task on a toy dataset such as MNIST. Results were verified using dimensionality reduction techniques like t-distributed Stochastic Neighbor Embedding (t-SNE) and Uniform Manifold Approximation and Projection (UMAP), alongside clustering algorithms such as k-means and Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN). ResultsThe VAEs encoder inherently maps data points into a latent space exhibiting discernible cluster structure, as evidenced by alignment with ground truth labels. While dimensionality reduction techniques (both t-SNE and UMAP) facilitated the application of clustering algorithms (k-means and HDBSCAN), these methods were primarily used to visualize and interpret the latent space organization. ConclusionsThis study demonstrates that VAEs effectively cluster data by implicitly encoding assignments in their latent representations. Determining cluster membership from encoder output, combined with reconstruction likelihood using semantic features, offers a principled approach for identifying typical samples and anomalies. Future research should focus on leveraging this inherent clustering capability of VAEs to enhance interpretability and facilitate clinical application.

Electroencephalography (EEG) interpretation in clinical practice relies on the analysis of energy distribution across standard frequency bands. The Weed Plot framework encodes band-wise spectral energy, computed using Fourier-based methods, into a symbolic representation that preserves the interpretability of traditional EEG analysis. In this study, we propose a wavelet-based extension of this framework, where the energy of predefined clinical EEG bands is estimated using the Discrete Wavelet Transform instead of Power Spectral Density. Unlike Fourier-based approaches, wavelets provide a time-frequency representation that captures transient and non-stationary dynamics while remaining consistent with clinically defined bands. From these estimates, symbolic patterns are constructed based on the relative ordering of frequency bands within short temporal windows. Their empirical distribution is used to extract entropy-based features for epilepsy detection using multiple machine learning classifiers. From an Artificial Intelligence perspective, the main contribution is a structured symbolic encoding that enhances feature discriminability. From an engineering perspective, the contribution lies in an automated framework for EEG-based epilepsy detection. Experimental results show that wavelet-based representations improve classification performance compared to raw entropy and Fourier-based features. This improvement arises from the interaction between time-frequency localization and symbolic encoding, producing more discriminative feature distributions. These findings support wavelet-based symbolic representations as a robust and interpretable framework for EEG analysis, bridging clinical interpretation and data-driven methods.

The opportunity to collect movement data from smartphones for prolonged periods has opened new perspectives in the field of clinical movement analysis. However, when monitoring peoples mobility in free-living conditions, smartphone placement can influence the validity of the extracted digital mobility outcome. This study aimed to develop and validate an automatic smartphone placement recognition classifier and to investigate potential critical factors that can influence performance. The classifier was trained on data from 15 healthy participants using inertial signals collected from smartphones placed at six body placements during free-living walking and externally validated on over 3,000 individuals from external datasets, including blind participants and patients with cardiovascular or Parkinsons disease. A decision-tree ensemble model was developed using feature subsets of increasing dimensionality, with the optimal subset comprising 50 features. Classification accuracy increased consistently when front and back pocket placements were aggregated (81.1%) and further improved when coat pocket was also included in the pocket class (88.5%), underscoring the challenge of distinguishing between fine-grained pocket placements. The best-recognized placements across the external datasets were lower back (precision: 100%, recall: 72.5%), hand (precision: 94.2%, recall: 94.5%), and the aggregated pocket class (precision: 86.7%, recall: 90.2%). Recognition accuracy changed across cohorts (0.73 - 0.85), activities (0.63 - 0.94) and speed (0.79 - 0.87), however it stayed consistent across various technological and environmental factors. Overall, this study demonstrates the feasibility of robust placement recognition in walking and underscores the importance of accounting for key influencing factors when designing frameworks intended for deployment in heterogeneous real-world or clinical contexts. HighlightsO_LIMachine learning accurately identifies smartphone placement during real-world gait C_LIO_LISix on-body placements recognized, including pockets, hand, bag, and lower-back C_LIO_LIFree-living data used for training, ensuring robust performance across conditions C_LIO_LIFeature selection and hyperparameter tuning optimize classification accuracy C_LIO_LIExternal validation confirms generalizability across >3,000 healthy and diseased adults C_LI

Physical activity (PA) plays an important role in maintaining and improving health. Daily steps have been a key PA measure that is easily accessible with common wearable devices. However, methods are lacking to recommend a personalized optimal distribution of daily steps over a period of time for the best of certain health biomarkers. In this paper, we fill this void based on the data from the All of Us Research Program which includes months of step counts as well as repeated measurements of key health biomarkers. We develop a new offline reinforcement learning (RL) algorithm to learn personalized and optimal PA distributions associated with cardiometabolic risk, where the action is a function representing the daily step distribution over a period of time. Simulation studies demonstrate the advantage of the proposed approach over existing continuous-action RL methods. The learned optimal policy from the All of Us data generally suggests people take more daily steps and also follow a more consistent pattern of PA over time while offering tailored recommendations for subgroups in blood glucose level, body mass index, blood pressure, age, and sex.

Whole-body SPECT bone scintigraphy reflects skeletal metabolic activity throughout the body and plays an indispensable role in the screening, treatment evaluation, and prognostic assessment of bone metastases in tumors. However, the automatic detection and segmentation of hypermetabolic bone lesions remain challenging due to low contrast, limited spatial resolution, and complex lesion distributions. In this study, we proposed Bone-Segnet, a dual-view guided automatic segmentation network for hypermetabolic bone lesions that integrated multi-scale feature modeling, global context modeling, and view-conditioned modulation. Pixel-level annotated anterior and posterior whole-body bone scintigraphy images were used for model training and prediction. The proposed network enhanced the recognition of low-contrast and small-scale lesions through small-lesion enhancement and multi-scale contextual modeling. A Transformer module was further introduced to strengthen global feature representation, while cross-view collaborative modeling was achieved by incorporating the complementary characteristics of anterior and posterior imaging. Experimental results demonstrated that the proposed method outperformed existing approaches across multiple evaluation metrics, with the Dice score improving from 0.7440 to 0.8750, indicating a substantial improvement in segmentation performance. Further quantitative analysis based on the segmentation results revealed significant differences among disease types in lesion count, pixel burden, and spatial distribution patterns, reflecting the heterogeneity of disease-related skeletal metabolic activity. Overall, the proposed method improved automatic lesion segmentation performance and enabled quantitative analysis of lesion burden and spatial distribution patterns, providing objective data support for the assessment of related diseases. Index Terms--Whole-body SPECT, bone lesion segmentation, dual-view modeling, quantitative analysis.

Ovarian cancer is one of the gynecological cancer types, which, if metastasized and not detected early, can cause deaths among women. Therefore, there is a need to accurately predict drug responses to ovarian cancer. A gynecological pathologist inspects abnormality in tissues, followed by providing a report about patients; however, such a diagnostic process is (1) hard; (2) requires experience; and (3) time consuming. Moreover, existing tools are far from perfect. Hence, we present a computational pipeline to improve predicting drug response pertaining to ovarian cancer, derived as follows. First, we download digital pathology images pertaining to ovarian bevacizumab response from the cancer imaging archive repository. We employed histogram of oriented gradients to images, constructing feature vectors, provided to Fisher linear discriminant analysis to change the representation through dimensionality reduction. Then, we provide reduced-dimensionality data for regression analysis through support vector regression coupled with various kernels and calculating the area under the ROC curve (AUC). Experimental results against transformer-based models (ViT and Swin) and other deep learning (DL) models (VGG16, ResNet50, InceptionV3, MobileNetV2, and EfficientNetB6) demonstrate that our approach with radial kernel (named SVRD+R) yielded an AUC performance improvements of 17% against the best-performing transformer-based model (ViT) while obtaining an AUC performance improvements of 14.9% when compared against the best DL-based model (MobileNetV2). These results demonstrate the superiority and feasibility of our AI-based pipeline when tackling prediction problems pertaining to gynecologic cancer studies. MSC92B05; 68T09

Millimeter-wave radar can quietly monitor health and behavior at home, which is vital for supporting people living with dementia. Most studies, however, remain limited to short-term testing in controlled spaces. Real-world deployment requires robust activity classification as a prerequisite: vital-sign and behavioral sensing require fundamentally different processing pipelines, and absent periods need to be reliably distinguished from stationary states. Bridging the critical gap between controlled laboratory demonstrations and continuous home monitoring, this paper introduces a self-adapting radar framework that extracts meaningful behavioral segments from massive, unconstrained real-world data. The system performs continuous real-time activity classification (stationary, walking, and absent) and target localization, selectively directing downstream processing to the most informative segments. It addresses key real-world deployment challenges including adaptive thresholding across subjects and environments, and walking detection under naturalistic activity conditions. Prior to integration with the Minder platform, the system was validated in a fully instrumented studio apartment against ground truth. Across 12 subjects, the system achieved an overall classification accuracy of 0.98, with F1 scores of 0.99 for absence and stationary states, and 0.95 for walking. Event-based evaluation yielded a per-subject walking sensitivity of 0.916{+/-} 0.058 and F1 score of 0.935 {+/-}0.030. Localization root mean square error during movement was 0.40 m. The results demonstrate reliable performance suitable for transitioning to long-term real-world home deployment.

Brain tumor detection using Magnetic Resonance Imaging (MRI) remains a challenging task due to tumor heterogeneity and imaging variability. This paper presents a novel hybrid Deep Convolutional Neural Network-Whale Optimization Algorithm (DCNN-WOA) framework for automated brain tumor detection and classification. The proposed method consists of four main stages: MRI data preprocessing and augmentation, deep feature extraction using multi-layer Convolutional Neural Networks (CNN), feature selection and hyperparameter optimization via the Whale Optimization Algorithm (WOA), and final classification with comprehensive performance evaluation. By jointly optimizing deep features and training parameters, the framework effectively reduces feature redundancy, accelerates convergence, and enhances model generalization. Experimental results on a publicly available MRI dataset demonstrate that the DCNN-WOA model outperforms conventional CNN and state-of-the-art Deep Learning (DL) architectures, achieving an accuracy of 97.8%, sensitivity of 96.4%, specificity of 98.1%, and F1-score of 97.2%. The practical impact of this approach makes it a promising solution for real-time clinical decision-support systems in neuroimaging.

Accurate pose estimation is central to automated infant General Movements Assessment during the fidgety period, when subtle limb movements, particularly at distal joints inform neurodevelopmental risks. Robust 2D pose tracking from handheld videos remains challenging in real-world settings, where occlusion, rapid motions, and visually ambiguous smaller joints frequently compromise anatomical accuracy. We present CRADLE, a clinically motivated, anatomy-aware post-processing pipeline designed to refine infant 2D movement trajectories across 24-anatomocal landmarks detected by our DeepLabCut-trained model. CRADLE integrates segment-length constraints, velocity-based anomaly detection, anatomically constrained interpolation, and Kalman filtering to correct both large localization failures and subtle persistent joint misplacements without relying primarily on confidence scores. Evaluations against conventional Confidence-Thresholding using Mean Absolute Error (MAE), {Delta}MAE, average Percentage of Correct Keypoints, and net keypoint correction rate showed consistently reduced or preserved error while maintaining accurate trajectories, with the strongest gains achieved at clinically important distal joints. Mean improvements reached up to 5 pixels for some smaller distal landmarks, large-magnitude corrections occurred more often than with Confidence-Thresholding, and well-localised joints remained largely unaffected. Positive net correction rates across metacarpophalangeal and metatarsophalangeal distal-landmarks further confirmed a favourable correction-degradation balance. By improving pose trajectory quality, CRADLE enhances the reliability of downstream movement analysis.

Electroencephalography (EEG) preprocessing is a critical yet time-consuming step that often relies on expert-driven, semi-automatic pipelines, limiting scalability and reproducibility across large datasets. In this work, we present sEEGnal, a fully automated and modular pipeline for EEG preprocessing designed to produce outputs comparable to expert-driven analyses while ensuring consistency and computational efficiency. The pipeline integrates three main modules: data standardization following the EEG extension of the Brain Imaging Data Structure (BIDS), bad channel detection, and artifact identification, combining physiologically grounded criteria with independent component analysis and ICLabel-based classification. Performance was evaluated against manual preprocessing performed by EEG experts at two complementary levels: preprocessing metadata (bad channels, artifact duration, and rejected components) and EEG-derived measures. In addition, test-retest analyses were conducted to assess the stability of the pipeline across repeated recordings. Results show that sEEGnal achieves performance comparable to expert-driven preprocessing while preserving key neurophysiological features. Furthermore, the pipeline demonstrates reduced variability and increased consistency compared to human experts. These findings support sEEGnal as a robust and scalable solution for automated EEG preprocessing in both research and large-scale applications. HighlightsFully automated and modular EEG preprocessing pipeline. Benchmarked against expert-driven preprocessing. Comparable performance in metadata and EEG-derived measures. Demonstrates stable performance in test-retest recordings. BIDS-based framework for reproducible EEG data handling.

Coronary artery disease is a leading cause of morbidity and mortality. Invasive coronary angiography is currently the gold standard in disease diagnosis. Several studies have attempted to use artificial intelligence (AI) to automate their interpretations with varying levels of success. However, most existing studies cannot generate detailed angiographic reports beyond simple classification or segmentation. This study aims to fine-tune and evaluate the performance of a Vision-Language Model (VLM) in coronary angiogram interpretation and report generation. Using twenty-thousand angiogram keyframes of 1987 patients collated across four unique datasets, we finetuned InternVL2-4B model with Low-Rank Adaptor weights that can perform stenosis detection, anatomy labelling, and report generation. The fine-tuned VLM achieved a precision of 0.56, recall of 0.64, and F1-score of 0.60 for stenosis detection. In anatomy segmentation, it attained a weighted precision of 0.50, recall of 0.43, and F1-score of 0.46, with higher scores in major vessel segments. Report generation integrating multiple angiographic projection views yielded an accuracy of 0.42, negative predictive value of 0.58 and specificity of 0.52. This study demonstrates the potential of using VLM to streamline angiogram interpretation to rapidly provide actionable information to guide management, support care in resource-limited settings, and audit the appropriateness of coronary interventions. AUTHOR SUMMARYCoronary artery disease has heavy disease burden worldwide and coronary angiogram is the gold standard imaging for its diagnosis. Interpreting these complex images and producing clinical reports require significant expertise and time. In this study, we fine-tuned and investigated an open-source VLM, InternVL2-4B, to interpret and report coronary angiogram images in key tasks including stenosis detection, anatomy identification, as well as full report generation. We also referenced the fine-tuned InternVL2-4B against state-of-the-art segmentation model, YOLOv8x, which was evaluated on the same test sets. We examined how machine learning metrics like the intersection over union score may not fully capture the clinical accuracy of model predictions and discussed the limitations of relying solely on these metrics for evaluating clinical AI systems. Although the model has not yet achieved expert-level interpretation, our results demonstrate the potential and feasibility of automating the reporting of coronary angiograms. Such systems could potentially assist cardiologists by improving reporting efficiency, highlightning lesions that may require review, and enabling automated calculations of clinical scores such as the SYNTAX score.

IntroductionTo propose an improved U-Net-based segmentation model for colorectal polyp segmentation, aiming to address the challenges of variable lesion morphology, ambiguous boundaries, complex background interference, and insufficient cross-level feature fusion in endoscopic images [5,12]. MethodsAn improved network termed MCA-UNet was developed based on U-Net [5]. The model incorporates a multi-scale context convolution block (MCCB) to enhance multi-scale feature extraction and an attention-guided feature fusion module (AGFF) to optimize skip-feature selection and fusion in the decoder. Experiments were conducted on publicly available colorectal polyp image datasets, including Kvasir-SEG and CVC-ClinicDB [13-15]. Four models, including U-Net, U-Net+MCCB, U-Net+AGFF, and MCA-UNet, were compared, and all models were trained for 100 epochs. Dice, intersection over union (IoU), and mean absolute error (MAE) were used as the main evaluation metrics [20]. ResultsOn the mixed validation set, the Dice scores of U-Net, U-Net+MCCB, U-Net+AGFF, and MCA-UNet were 0.742, 0.771, 0.754, and 0.783, respectively; the corresponding IoU values were 0.603, 0.635, 0.618, and 0.649; and the MAE values were 0.102, 0.090, 0.097, and 0.086. Compared with the baseline U-Net, MCA-UNet improved Dice and IoU by 5.53% and 7.63%, respectively, while reducing MAE by 15.69%. Comparisons on the Kvasir-SEG and CVC-ClinicDB validation subsets further demonstrated the more stable performance of the proposed model. ConclusionBy jointly integrating multi-scale contextual modeling and attention-guided feature fusion, MCA-UNet effectively improves the accuracy and robustness of colorectal polyp segmentation and may provide useful support for intelligent endoscopic image analysis [12,17,18].

Goal-oriented semantic communication has recently emerged in wireless sensor-actuator networks, emphasizing the meaning and relevance of information over raw data delivery, thereby enabling resource-efficient telecommunication. This paradigm offers significant benefits for intra-body or implantable sensor-actuator networks, including dramatic reductions in bandwidth requirements, latency, and power consumption. In this paper, we address a patch-based energy-efficient anomaly detection method for smart capsule endoscopy. We propose a deep learningbased algorithm that employs the similarity between features extracted from measured images and a reference (normal) image as the detection metric. The algorithm is evaluated using a clinical dataset of capsule-captured images, combined with a simulated intra-body channel model. The results demonstrate that even with only 60% of the transmission power (relative to a standard link design for QPSK modulation) and 65% of the light intensity, the probability of anomaly detection remains above 85%, and it gradually improves as power and illumination levels increase. This improvement translates into a potential battery life extension of over 43%. The findings highlight the potential of semanticaware, energy-efficient intra-body devices for more sustainable and effective medical interventions.

Falls among older adults can result in hip fractures that requires x-ray based assessment at emergency department (ED). Only 25.7% of patients presenting to EDs are diagnosed with a hip fracture, as such improved diagnosis prior to transportation to hospital could result in fewer hospital visits and improved triaging. Patient with hip fracture could be immediately directed to centres with orthopaedic surgeons, allowing for reduced time-to-surgery, particularly in rural communities. Ultrasound (US) imaging is portable and can identify fractures but requires expertise, particularly related to image interpretation. Deep learning may reduce operator dependence by automating image interpretation. This study aims to develop HipSAFE, a hip fracture detection tool on US, to support triaging by nurses and paramedics. We hypothesize that diagnostic accuracy will be comparable to pelvic x-ray diagnostic performance in a preclinical study. Bilateral hind limbs of 15 porcine cadavers were imaged by US-naive operators before and after an iatrogenic hip fracture. The limbs were divided into training, validation, and test (8 femurs) sets. The training data were augmented (geometric and photometric transformations). The models included MobileNetV3 (S/L), EfficientNet-Lite (0-2), and ResNet (18/50). Using a moving average aggregation on the operator cine clips, EfficientNet-Lite0 achieved the highest performance (F1=0.944 [95% CI:0.880-0.987]; sensitivity=89.5% [78.6-97.5%]; specificity = 100.0% [100.0-100.0]). The majority voting ensemble model ranked second (F1=0.932 [0.857-0.984]). Naive operators and radiologists had lower performance (F1=0.667 [0.596-0.758] and 0.685 [0.597-0.729]). This pre-clinical study demonstrated that HipSAFE has excellent diagnostic accuracy and there may be a role for US in improving hip trauma triaging, especially for rural and resource-constrained environments.

Machine learning models have seen a lot of success in medical image segmentation domain. However, one of the challenges that they face are confounders or shortcuts: spurious correlations or biases in the training data that affect the resulting models. One example of such confounders for surgical machine learning is the setup of surgical equipment, including tools and lighting. Using the task of identification of safe and dangerous zones of dissection in laparoscopic cholecystectomy images and videos as a use-case, we inspect two equipment-induced biases: the presence of surgical tools in the field of view and the position of lighting. We propose methods for evaluating the severity of these biases and augmentation-based methods for mitigating them. We show that our tool bias mitigations improve the models consistency under tool movements by 9 percentage points in the most inconsistent cases, and by 4 percentage points on average. Our lighting bias mitigations help reduce fraction of true dangerous zone pixels that may be predicted as safe under light changes from 5% to 1.5%, without compromising segmentation quality.

Optimizing assistive wearable devices is crucial for their efficacy and user adoption, yet state-of-the-art methods like Human-in-the-Loop Optimization (HILO) and biomechanical modeling face limitations. HILO is time-consuming and often restricted to optimizing control parameters, while inverse dynamics assumes invariant kinematics, which is unreliable for adaptive human-device interaction. Predictive simulation offers a powerful alternative, enabling computational exploration of design spaces. However, existing approaches often lack systematic optimization frameworks and rigorous validation against experimental data. To address this, we developed a Design Optimization Platform that integrates predictive simulations within a two-level optimization structure for personalizing assistive device design. This paper primarily validates the platforms predictive simulations against a publicly available dataset of the passive Biarticular Thigh Exosuit (BATEX), assessing its reliability. Our findings show that the model can sufficiently predict the kinematics and major muscle activations, except for the pelvis tilt and some biarticular muscles. The key finding is that successful identification of personalized optimal BATEX stiffness parameters needs acceptable prediction of metabolic cost trends, not their precise values. Our analysis further reveals that the models accuracy in predicting Vasti muscle activation in the baseline condition is a significant indicator of its success in predicting metabolic cost trends. This demonstrates that accurate prediction of performance trends is more important for effective simulation-based design optimization than perfect biomechanical accuracy, advancing targeted and efficient assistive device development.

Reliable interpretation of electrocardiograms (ECGs) requires precise identification of P, QRS, and T (PQRST) wave boundaries. However, it remains challenging due to noise, signal quality variability, and inherent morphological diversity particularly in recordings from children. This study systematically compares the performance of leading deep neural networks (DNN) and heuristic-based delineation algorithms on ambulatory single-lead ECG signals focusing on temporal accuracy. Experiments were conducted using the publicly available LUDB dataset and a private validation dataset comprising 21,759 annotated single-lead wave segments from 611 children recorded using KardiaMobile ECG sensor. DNN were first trained on the LUDB dataset and subsequently tested on the validation dataset. The delineation performance was assessed using Sensitivity (Se) and positive-predictive-value (P+) metrics. The best-performing heuristic based and DNN models reached Se and P+ of (98.9% vs 97.9%) for P, (99.8% vs 99.2%) for QRS, and (98.7% vs 95.9%) for T wave fiducials, respectively. The lowest standard-deviation (in ms) of wave onset/offset delineation was achieved by attention based 1DU-Net model; {+/-}16.6/{+/-}16.3 for P-wave, {+/-}14.0/{+/-}16.3 for QRS, and {+/-}26.3/{+/-}18.8 for T-wave, respectively. The findings indicate that optimized heuristic models can perform comparably to complex DNN, highlighting their efficiency and suitability for real-time ECG delineation in digital health monitoring applications.