Biomedical Signal Processing and Control

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.

Background: Atrial fibrillation (AFib) is the most common sustained arrhythmia in the world, imposing a heavy clinical and economic burden on global healthcare systems. Early detection of AFib can reduce mortality and morbidity, while helping to alleviate the growing economic burden of cardiovascular diseases. With the increasing availability of digital health technologies, computational solutions have great potential to support the timely diagnosis of cardiac abnormalities. Objectives: With the increasing availability of electrocardiogram (ECG) data from clinical and wearable devices, manual interpretation has become impractical due to its time-consuming and subjective nature. Existing automated approaches often rely on single classifiers or fixed ensembles that primarily optimize predictive accuracy while neglecting model diversity, which leads to limited robustness and generalization across heterogeneous datasets. Therefore, this study aims to develop a robust and diversity-aware framework for automatic AFib detection that simultaneously improves classification performance and model generalizability. To this end, we propose MOE-ECG, a multi-objective ensemble selection and fusion framework that explicitly optimizes both predictive performance and inter-model diversity for reliable AFib detection from ECG recordings. Methods: The proposed multi-objective ensemble (MOE) framework uses ensemble selection as a bi-objective optimization problem and employs multi-objective particle swarm optimization to identify complementary classifiers from a heterogeneous model pool. Unlike conventional ensembles, it explicitly optimizes both predictive performance and diversity and integrates Dempster-Shafer theory for uncertainty-aware decision fusion. After filtering the ECG signals to remove baseline wander and noise, they were segmented into windows of 20, 60, and 120 heartbeats with 50% overlap. The proposed approach was evaluated over five independent runs to assess its stability and generalization. Fifteen statistical and nonlinear features were obtained from the RR-intervals of the pre-processed ECG signals, of which eight features were selected with correlation analysis to capture subtle information from the ECG data. We trained and evaluated the performance of the proposed model in three open source databases, namely, the MIT-BIH Atrial Fibrillation Database, Saitama Heart Database Atrial Fibrillation, and Long-Term AF Database. Results: The proposed approach achieved the best overall performance on 60-beat segments, with an average accuracy of 89.85%, precision of 91.14%, recall of 94.19%, an F1-score of 92.64%, and area under the curve (AUC) of around 0.95. Statistical analysis using Holm-adjusted Wilcoxon tests confirmed significant improvements (p<0.05) compared to both the best individual classifier and the unoptimized average ensemble of all classifiers. These findings show that the proposed selection and evaluation methodology, rather than group aggregation alone, is the key driver of performance improvements. Conclusion: The results obtained demonstrate that the MOE-ECG model offers a robust, accurate, and reliable solution for the detection of AFib from short ECG segments. The empirical findings, in general, confirm that multi-objective ensemble fusion enhances diagnostic performance and offers robust predictions that will open up possibilities for real-time AFib detection in clinical and tele-health settings.

Brain tumors are one of the most life-threatening diseases, requiring precise and timely detection for effective treatment. Traditional methods for brain tumor detection rely heavily on manual analysis of MRI scans, which is time-consuming, subjective, and prone to human error. With advancements in deep learning, Convolutional Neural Networks (CNNs) have become popular for medical image analysis. However, CNNs are limited in their ability to capture spatial hierarchies and pose variations, which reduces their accuracy, particularly for tasks like brain tumor segmentation where precise spatial relationships are crucial. This research introduces a hybrid Capsule Neural Network (CapsNet) and ResNet50 model designed to overcome the limitations of traditional CNNs by capturing both spatial and pose information in MRI scans. The proposed model leverages ResNet50 for feature extraction and CapsNet for handling spatial relationships, leading to more accurate segmentation. The study evaluates the model on the BraTS2020 dataset and compares its performance to state-of-the-art CNN architectures, including U-Net and pure CNN models. The hybrid model, featuring a custom 5-cycle dynamic routing algorithm to enhance capsule agreement for tumor boundaries, achieved 98% accuracy and an F1-score of 0.87, demonstrating superior performance in detecting and segmenting brain tumors. This study pioneers the systematic evaluation of the ResNet50 + CapsNet hybrid on the BraTS2020 dataset, with a tailored class weighting scheme addressing class imbalance, improving effectiveness in identifying irregularly shaped tumors and smaller regions in identifying irregularly shaped tumors and smaller tumor regions. The study offers a robust solution for automating brain tumor detection. Future work will explore the use of Capsule Networks alone for brain tumor detection in MRI data and investigate alternative Capsule Network architectures, as well as their integration into clinical decision support systems.

Accurate diagnosis of cardiac abnormalities from electrocardiogram signals remains a central challenge in automated cardiovascular assessment. This study investigates the efficiency of time-frequency representations and deep learning architectures in classifying 12-lead ECGs into five diagnostic super-classes using the PTB-XL dataset. Continuous Wavelet Transform is applied to generate time- frequency representations, scalograms and phasograms, representing spectral energy and phase distributions, respectively. We experiment with both early and late information fusion strategies using several convolutional and transformer-based networks of a custom Convolutional Neural Network, Hybrid Deep Learning, transfer learning, feature fusion, and ensemble modeling, and weighted loss strategies. An ensemble fusion of models trained on time-frequency representation and time representation achieved the best overall performance of Area Under Curve of 0.9233 surpassing individual modalities. To improve the results further, weighted focal loss is used to improve the low classification rates in some labels due to imbalanced data. The results highlight the potential of multi-representation wavelet fusion for interpretable and generalizable ECG classification.

Accurate detection of interictal epileptiform discharges (IEDs) in electroencephalography (EEG) plays a crucial role in epilepsy diagnosis. Our work investigates the classification of IEDs using Artificial Neural Networks (ANNs) trained on EEG data represented in both signal and source space. Source waveforms were computed using equivalent current dipole models fitted using either a 1-parameter fixed-orientation or a 3-parameter projection approach, both localized to a single best-fit position during the rising flank of the IED. The ANN was trained on raw and feature-extracted versions of signal space and source space data. Feature extraction significantly improved performance across all domains. The highest accuracy (0.98) was achieved in signal space using Katz Fractional Dimension (KFD). In source space analyses, the 1-parameter and 3-parameter models achieved a maximum accuracy of 0.84, with statistical features performing best for the fixed-orientation model and KFD for the free orientation model. Additionally, annotations from three independent expert markers showed considerable variability, with ANN performance falling within the range of inter-expert agreement. These findings support the potential of ANN-based tools to assist expert evaluation in future clinical workflows.

We present an EEG-based approach to characterize disease-related spectro-temporal signatures in Alzheimers disease (AD) and Parkinsons disease (PD). To this end, key spectral features were first identified using explainable machine learning, and their temporal dynamics were then examined to characterize variability patterns and statistical properties. EEG recordings were segmented into non-overlapping 4-s epochs, from which spectral features based on relative band power and spectral entropy were extracted. Random Forest classifiers were trained to discriminate individual subjects with AD and PD from healthy controls (HC) using a Leave-One-Subject-Out Cross-Validation (LOSOCV) strategy. The most discriminative spectral features and the directionality of their contributions were identified through a SHAP-based explainable analysis. Subsequently, the temporal dynamics of the key features were analyzed to characterize disease fingerprints in terms of variability at both inter-subject and intra-subject levels and their distributional profiles. Our results confirmed spectral slowing in both disorders and revealed disorder-specific differences in the dominant spectral markers: the theta/alpha ratio was the most influential feature for AD, whereas mean relative theta power was the primary feature for PD discrimination. We show that increased variability in key spectral features is a distinguishing signature of AD and PD, with disease groups exhibiting greater inter-subject heterogeneity and higher intra-subject temporal variability than HC. Moreover, the key features showed heavy-tailed behavior, for which a lognormal model provided a plausible fit across groups. We conclude that this EEG-based characterization provides a meaningful avenue for tracking deviations from healthy neural activity.

Sustained attention is an important requirement for high performance in all cognitive processes. Quantifying the level of sustained attention to prevent attention lapses is therefore necessary for effective human-machine interfacing. Furthermore, sustained attention evaluation can help diagnose and treat attention deficit hyperactivity disorders. Attention level can be assessed by brain and heart signals. This study employed functional near infrared spectroscopy (fNIRS) and the heart rate variability (HRV) information extracted from the fNIRS signals to differentiate the rest and three levels of sustained attention states. Sustained attention states are induced by three modified versions of continuous performance tests (CPT). Eight subjects engaged in three sessions of attention tests. fNIRS brain signals were recorded from the right prefrontal and dorsolateral prefrontal cortex. HRV information was then extracted by processing the fNIRS signals. For attention classification, support vector machine (SVM), linear discriminant analysis (LDA), and random forest (RF) algorithms with mutual information based feature selection were applied on the fNIRS and HRV data both separately and together. In the classification of the three levels of attention using fNIRS and HRV data, the LDA classifier showed the best performance accuracy of (80.9 {+/-} 1.5%) and (56.2 {+/-} 1.0%), respectively. For two-class classification between the rest and the attention states (all together), the accuracies of (98.9 {+/-} 0.3%), (95.6 {+/-} 1.2%), and (99.5 {+/-} 0.2%) were obtained using the RF classifier on the fNIRS, HRV, and combined data, respectively. These results demonstrate the effectiveness of the HRV data for classifying sustained attention states. Moreover, using the combined fNIRS and HRV data provides better classification accuracy.

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.

I.AO_SCPLOWBSTRACTC_SCPLOWCoronary Artery Disease (CAD) is a leading cause of cardiovascular-related mortality and affects 20.5 million people in the United States and approximately 315 million people worldwide in 2022. The asymptomatic and progressive nature of CAD presents challenges for early diagnosis and timely intervention. Traditional diagnostic methods such angiography and stress tests are known to be resource-intensive and prone to human error. This calls for a need for automated and time-effective detection methods. In this paper, this paper introduces a novel approach to the diagnosis of CAD based on a Convolutional Neural Network (CNN) with a temporal attention mechanism. The model will be developed on an architecture that will automatically extract and emphasize critical features from sequential medical imaging data from coronary angiograms, allowing subtle signs of CAD to be easily spotted, which could not have been detected by convention. The temporal attention mechanism strengthens the ability of a model to focus on relevant temporal patterns, thus improving sensitivity and robustness in detecting CAD for various stages of the disease. Experimental validation on a large and diverse dataset demonstrates the efficacy of the proposed method, with significant improvements in both detection accuracy and processing time compared to traditional CNN architectures. The results of this study propose a scalable solution system for the diagnosis of CAD. This proposed system can be integrated into clinical workflows to assist healthcare professionals. Ultimately, this research contributes to the field of AI-driven healthcare solutions and has the potential to reduce the global burden of CAD through early automated detection.

Developing a reliable detection of olfactory performance for early Alzheimers disease (AD) diagnosis remains challenging. Existing methods, such as psychophysical and event-related potential approaches, provide limited consistency in quantifying olfactory function. This study introduces a novel and objective framework that analyzes olfactory-stimulus-evoked EEG synchronizations of the subjects for AD diagnosis. We calculated the time-resolved wavelet coherence between EEG signals and then determined the timings (i.e., latency and duration) that describe when olfactory-stimulus-induced EEG channel interactions begin and end for each channel and frequency band. These timings, as well as the mean synchronization values in these segments, were used as features for diagnosis. Our framework, when cross-correntropy was used as a synchronization measure, exhibited a notable diagnostic accuracy in mild AD detection. The most discriminating feature between mild AD and healthy subjects was found to be the latency of synchronization between Fp1 and Fz in the low{theta} band, which showed significantly high correlation with clinical test scores. Furthermore, our framework achieved 100% diagnosis accuracy when EEG features and clinical test scores were used together. Our findings show that inter-channel short-lived synchronization timings serve as useful and complementary metrics about subjects olfactory performance and their neurological conditions.

This research presents multimodal deep learning for structural heart disease prediction. We evaluated multiple deep learning architectures, including TCN, Simple CNN, ResNet1d18, Light transformer and Hybrid model. The models were examined across the three seeds to ensure robustness, and bootstrap confidence interval is used to measure performance differences. TCN consistently outperforms other competing architectures, achieving statistically significant improvements with stable performance across runs. Similarly in predictive analysis, TCN has efficient computation and stable training compared to all competing architectures. Our results show that TCN emphasizes fairness evaluation when developing deep learning models for healthcare applications.

AO_SCPLOWBSTRACTC_SCPLOWDecoding imagined speech from electroencephalography (EEG) recordings is potentially useful for brain-computer interfaces. Previous studies have focused on decoding semantic information from EEG, leaving the decoding of emotion - an important component of human communication - largely unexplored. Here, we report two experiments involving participants tasked with overt (n = 14) or imagined (n = 21) emotional vocalisation in five different categories: anger, happiness, neutral, sadness, and pleasure. Throughout, we recorded 64-channel EEG; we computed time-frequency features and used a logistic-regression classifier to evaluate emotion decoding accuracy. In five participants, we also recorded facial surface electromyography (sEMG) during imagined vocalisation, and studied the contamination of EEG by sEMG. Our results show that emotion can be decoded from single-trial EEG recordings of both overt (78.1%, chance = 20%) and imagined vocalisation (36.4%). The high-gamma band (50 to 100 Hz) and lateral EEG channels (T7, T8, and proximal) were important for decoding. sEMG analysis indicated that involuntary facial muscle activity contributed to these spectral and spatial patterns during imagined vocalisation, especially during happy vocalisations. We conclude that involuntary facial muscle activity is associated with certain emotion categories (i.e., happiness), and drives above-chance decoding of emotion from single-trial EEG recordings of imagined vocalisation.

Early and accurate diagnosis of Alzheimers Disease (AD) is critical for effective intervention. While previous studies have explored speech-based biomarkers for AD, this paper presents the first systematic investigation of acoustic, prosodic, and phonological speech features for detecting this neurocognitive disorder. Our study has two main objectives: (1) to assess the individual impact of AD on a wide range of speech features, and (2) to identify the most informative feature subsets using a combination of four feature ranking methods and seven machine learning classifiers. We conducted our experiments using the publicly available ADReSS Challenge dataset, allowing for direct comparison with prior speech-based approaches. Our analysis focused on continuous acoustic and prosodic measures as well as discrete phonological features, both independently and in combination. The best performance, with an F1-score of 0.89, was achieved using the optimal subset of acoustic features with ensemble learning and by integrating all three feature types, suggesting that combining continuous and categorical speech features offers complementary diagnostic value. This approach surpasses all previous speech-only methods in the ADReSS Challenge and comes close to the best reported overall result from the campaigns dataset, even without using lexical information. The results were further validated through positive outcomes on the Delaware corpus (focused on MCI) and a set of speeches from President Ronald Reagan (who was diagnosed with Alzheimers). These findings suggest that speech patterns, beyond just the content, are more indicative of Alzheimers disease than previously thought, underscoring the potential of multi-layered speech analysis for non-invasive AD detection.

The main challenges in the life of a child with autism are difficulties in communication, behavior, and social interaction. Early diagnosis of this neurodevelopmental disorder improves patient outcomes by enabling more effective, personalized interventions. This diagnosis can sometimes be difficult, especially in very young children. Non-invasive, relatively accessible, and able to reflect neural function in real time, electroencephalography (EEG) shows promise in the detection of Autism spectrum disorders (ASD). However, because EEG data is still difficult for experts to understand, machine learning and artificial intelligence (AI) are beginning to be used in this field as well. In this paper, a ResNet+BiLSTM hybrid deep network was applied and achieved high accuracy in distinguishing individuals with autism from neurotypical subjects. Since AI models typically provide predictions without clear explanations, this study employs explainable AI (XAI) methods such as SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) to clarify their decision-making.Delta, theta, alpha, beta, and gamma waves, as well as ERP components P100, N100, P200, MMN, and P600, were analyzed in the two neurotypical and autistic groups that were compared in this study using EEG recordings. By integrating SHAP and LIME, the system achieved both accurate classification and transparent explanations, pointing to EEG- and ERP-based features as reliable biomarkers for ASD.

Parkinsons disease is a progressive neurological disorder characterized by motor impairments whose severity is commonly assessed using the Unified Parkinsons Disease Rating Scale (UPDRS). Although clinically established, UPDRS assessment requires in-person evaluation by trained specialists and is inherently subjective, limiting its suitability for frequent monitoring. Speech production is affected early in Parkinsons disease and provides a non-invasive modality for remote symptom assessment. In this study, an uncertainty-aware personalized framework is proposed for estimating Parkinsons disease severity from speech signals. The approach integrates longitudinal temporal modeling of longitudinal speech recordings with patient-specific representations and a probabilistic latent disease state. Continuous motor UPDRS scores are estimated jointly with ordinal disease severity stages, enabling both fine-grained regression and clinically interpretable stratification. Predictive uncertainty is explicitly quantified, yielding confidence-aware severity estimates suitable for telemonitoring applications. The method is evaluated on a longitudinal speech dataset using a strict patient-wise split, ensuring that all test subjects are unseen during training. On the held-out test set, the proposed model achieves high predictive accuracy (mean absolute error 0.56 UPDRS points, root mean squared error 0.74, and coefficient of determination R2 = 0.99) for motor UPDRS estimation. Ordinal severity classification attains an accuracy of 0.92 across mild, moderate, and severe disease stages. Comparative experiments against classical machine learning methods and global temporal baselines demonstrate consistent performance improvements.These results indicate that personalized, uncertainty-aware modeling of speech signals can support accurate and clinically meaningful remote monitoring of Parkinsons disease severity.

Rapid and accurate detection of electrographic seizures is critical for both clinical diagnosis and neuroscience research. Although seizure identification is commonly performed in the time domain, analysis in the time-frequency domain provides a more comprehensive representation of seizure characteristics. In this study, we present a 3D convolutional neural network (CNN) that incorporates a trainable continuous wavelet transform (CWT) layer, enabling adaptive time-frequency feature learning directly from raw EEG. To address common data challenges, we augment the 3D CNN for pre-training with contrastive learning, comparing contrastive predictive coding (CPC) against bidirectional contrastive learning (BiCL). On single-channel and multi-channel data, the standard 3D CNN outperformed both a 2D CNN with pre-computed CWT and a 1D CNN that processes raw signals, achieving >95% accuracy down to 0.5-second segments. Compared to the standard 3D CNN, the 3D CNN with BiCL pre-training showed superior performance in both low-data and class imbalance scenarios. Further experiments involving band-pass filtering and temporal shuffling revealed that classification is driven primarily by low-frequency patterns and statistical features rather than temporal dependencies. The proposed framework also maintained >90% accuracy with moderate noise and downsampling applied to inputs, as well as when cross-subject generalization was evaluated using held-out subjects. We show that a 3D CNN with a trainable CWT layer and BiCL pre-training enables accurate sub-second seizure detection and effectively mitigates data limitations common in clinical settings. This work demonstrates that time-frequency embedding within CNNs, augmented by self-supervised pre-training, offers a promising path toward architectures for sub-second seizure detection in the presence of practical limitations of real-world scenarios.

ObjectiveWhile there are numerous reports on heart rate and its variabilities, a detailed analysis of the component intervals for healthy adults in well controlled condition is lacking. This study analyzes the effect of age, sex, and Body Mass Index (BMI) on nine resting electrocardiogram (ECG) intervals and their intra-individual variabilities in healthy adults under the same testing environment. MethodsUsing the "Autonomic Aging" dataset, ECG recordings from 1,121 healthy volunteers (ages 18-92) were processed. The study employed a specialized segmentation algorithm to identify key ECG markers. We analyze statistically how age, BMI, and sex impact the durations and variabilities of nine ECG intervals. ResultsFifty years of age serves as a critical transition age for cardiac aging for all subjects as a whole. Above this age, the active interval, which is the combined atrial and ventricular conduction time, increases three times faster than at a younger age, primarily driven by lengthening of depolarization times. Compared to the opposite sex, older low-BMI males have a longer atrial conduction time, and older low-BMI females have a larger variability in the ventricular conduction time. High BMI increases the heart rate by reducing the length of the idle interval, i.e., the isoelectric segment at the end of a cardiac cycle. The rate increase is more pronounced among older subjects than younger ones. High BMI males start to exhibit an elevated heart rate and larger variability in the atrial conduction time in their 30s. High BMI females start to show a larger variability in the ventricular repolarization time around 50 years old. ConclusionAge, BMI, and sex all have major impacts on the ECG intervals and their variability. A resting heart behaves largely like a pulse width modulation system, with a stable active interval and an adjustable idle interval to meet the varying needs for cardiac output. The durations and variabilities of the active interval, more than those of the RR interval, are indicators of a hearts health condition. A young and healthy heart tends to have a shorter duration and smaller variability in the active interval.

Electroencephalography (EEG) serves as a fundamental tool in modern neurology, cognitive neuroscience, and brain-computer interfaces, but its practical application is often compromised by artifacts. Physiological artifacts are particularly intractable due to overlapping spectral features with neural signals, hindering reliable EEG interpretation. In this work, we propose Grid-based 3D Convolution-Transformer (G3DCT), an interpretable deep learning framework for EEG artifact identification. The framework embeds multi-channel EEG signals into fixed grids to leverage electrode spatial topology, employs parallel multi-branch temporal convolutions and Transformers to handle complex artifacts, and incorporates an attention module to visualize scalp activation patterns, which enhances physiological interpretability. Our evaluation on three datasets demonstrates that G3DCT outperforms existing state-of-the-art models. For challenging combined artifacts, it secures a gain of 2.8% in F1-score over the second-best model. These results demonstrate that G3DCT provides an efficient and robust solution for EEG artifact identification, which has the potential to enhance the reliability of EEG-based applications in practice.

Blood pressure (BP) measurement is crucial for medical care, yet existing BP methods are either invasive, tethered, or suffer from low temporal resolution. Non-invasive continuous BP estimation thus remains a significant challenge. To address these challenges, this work presents a novel, non-invasive, multi-modal sensor designed for continuous blood pressure estimation using multiple biosignal modalities as feature inputs. From these input data, we extract cardiovascular timing intervals (e.g., pulse arrival time), which serve as key features for BP regression models, enabling continuous, non-invasive BP monitoring. We validate our algorithm with 16 healthy subjects using standard blood pressure cuff readings as ground truth. Our wearable, non-invasive multimodal and multinodal sensor array for integrated computation (MOSAIC) demonstrated promising performance and was able to predict systolic and diastolic BP across all study subjects with a MAE of 5.31 {+/-} 7.32 mmHg and 4.27 {+/-} 2.35 mmHg, respectively.

Handheld ultrasound devices have revolutionized point-of-care diagnostics, but their effectiveness remains limited by operator dependency and the need for specialized training. This paper presents an intelligent guidance and diagnostic assistance system for the handheld wireless ultrasound device, enabling automated carotid artery and thyroid examinations through handheld operation. Drawing inspiration from the Actor-Critic framework, we implement a simulation-based reinforcement learning approach for real-time probe navigation toward standard anatomical views. The system integrates YOLOv8n-based detection networks for carotid plaque and thyroid nodule identification, achieving real-time inference at 30 frames per second. Furthermore, we propose a hybrid measurement approach combining UNet segmentation with the Snake algorithm for precise biometric quantification, including carotid intima-media thickness (IMT), lumen diameter, and lesion dimensions. Experimental validation on clinical datasets demonstrates that the proposed system achieves 91.2% accuracy in standard plane acquisition, 87.5% mean average precision (mAP) for plaque detection, and 89.3% mAP for nodule identification. Measurement results show excellent agreement with expert sonographers, with IMT measurements exhibiting a mean absolute difference of 0.08 mm. These findings demonstrate the feasibility of intelligent handheld ultrasound examination, significantly reducing operator dependency while maintaining diagnostic accuracy comparable to experienced clinicians.