Journal of Electronics, Electromedical Engineering, and Medical Informatics

Machine Learning Approach to ATPG in VLSI Circuits Using Random Forest Regressor

2026-10-03T17:22:21+07:00

As the need for highly advanced chip technology and high-performance computing grows, semiconductor technology must develop rapidly. Also, as VLSI circuits become more complex, it is more difficult and time-consuming to test circuits using traditional test methods. Hence, Chip design, verification, and testing become extremely complex and critical processes. To confirm the chip as fault-free, testing examines its functionality, timing, and connectivity. Various automatic test pattern generation (ATPG) tools have been available for a long time for fault detection. For large combinational circuits, a lot of computing resources and more time are required to use such tools. To tackle this challenge, machine learning (ML) techniques have recently been introduced as a promising alternative to improve the efficiency of test generation. Machine learning has advanced dramatically over the last few years, and it currently plays an important role in enhancing automation, efficiency, and decision-making in a variety of domains. The main purpose of this work is to explore the possibility of using machine learning to predict test vectors in digital combinational circuits and achieve comparable fault coverage. The Random Forest Regressor method is utilized for generating test vectors. Experimental results showcase that this method performs well compared to traditional tools like ATALANTA on most ISCAS85 benchmark circuits. This approach reduced test pattern generation time to 0.0146 sec, while the standard tool requires 0.02 sec for test set generation, retaining almost similar fault coverage. This research demonstrated the highest fault coverage of 96.523 for the c5315_13 circuit and the lowest coverage of 60.404 for c1908_23, compared to the ATALANTA Tool. The results demonstrate that machine learning is a promising complementary testing method for conventional ATPG tools in VLSI testing.

STFT-Based Multiclass Heart Sound Classification Using BiLSTM and CNN-BiLSTM Models

2026-08-03T20:45:37+07:00

Cardiovascular diseases require early and reliable screening because manual auscultation may be affected by noise, subjective interpretation, and inter-observer variability. This study aimed to develop an STFT-based deep learning framework for multiclass phonocardiogram (PCG) classification. The proposed framework was designed to provide a reproducible evaluation procedure by combining standardized preprocessing, time–frequency feature extraction, and deep learning-based classification under the same experimental conditions. Unlike approaches that may evaluate segmented signals without clearly preserving recording-level separation, this study emphasizes a leakage-free splitting strategy to reduce the risk of overestimated performance and to provide a more reliable assessment of model generalization. The Yaseen PCG dataset, consisting of 1000 recordings from five classes (AS, MR, MS, MVP, and Normal), was divided using a leakage-free recording-level split before segmentation and spectrogram generation. After preprocessing, 2-second PCG segments with 50% overlap were converted into 128 × 128 STFT spectrograms and classified using BiLSTM and CNN-BiLSTM models. Both models were trained and tested using the same dataset split, preprocessing pipeline, and evaluation metrics, including accuracy, precision, recall, F1-score, specificity, and confusion matrices. The BiLSTM model achieved 92.36% accuracy in the final independent test run, while the CNN-BiLSTM model achieved 95.83%. Across three repeated runs, BiLSTM achieved 93.85% ± 1.47%, whereas CNN-BiLSTM achieved 95.94% ± 0.71%. These results show that CNN-BiLSTM provides higher and more stable classification performance for five-class PCG classification, while BiLSTM remains a simpler alternative for lightweight implementation. Overall, the proposed STFT-based framework provides a reliable approach for automated heart sound classification and may support future computer-aided cardiac screening applications.

Electrooculography-Based Voluntary Eye-Blink Detection for Arabic Assistive Communication

2026-08-11T18:14:56+07:00

People with profound neuromotor disabilities commonly experience significant speech disability and lack of voluntary control over their limbs, retaining, however, the ability to perform deliberate eyelid blinks. Such retained capability presents an opportunity to be used as assistive communication. Nevertheless, reliable detection of voluntary blinks and decoding of communication patterns based on them is challenged by variation in the waveforms, amplitudes, durations, and inter-blink intervals of the voluntary eye blinks. In this work, a cost-efficient vertical electrooculography (EOG) approach for detecting voluntary eye blinks and decoding predefined assistive messages in Arabic was presented and assessed. The pipeline consists of signal conditioning, root-mean-square envelope calculation, adaptive hysteresis-based thresholding, temporal segmentation, and rule-based interpretation of single- and double-blink patterns. Quiet-background threshold updating and locking it during the communication window period were used to increase the robustness of the pattern detection process, while autocorrelation analysis and temporal constraints were utilized to interpret patterns. A single blink is associated with binary 0, and a double blink with binary 1; four consecutive pattern positions form a 4-bit frame that represents one of sixteen predefined Arabic messages. Unclear patterns are excluded from processing and not mapped into any valid message. The proposed approach was evaluated offline on 4,000 word-level recordings from 25 healthy participants, with verification using videos. The F1-score of 95.69%, macro-F1-score of 98.30%, and message-decoding accuracy of 92.98% were obtained. These findings demonstrate the possibility of implementing an interpretable and cost-efficient vertical-EOG approach for message-level Arabic assistive communication without morse-code-type encoding and letter-by-letter spelling. Direct mapping of patterns into messages can help reduce communication effort and time. Nevertheless, since the present evaluation was conducted offline and included only healthy subjects, further validation with target users and real-time implementation should be conducted before deployment.

ProtoSurv-X: Explainable Brain Tumor Survival Prediction

2026-10-04T09:53:37+07:00

Accurate survival prediction for patients with high-grade glioma is important for prognostic assessment and personalized treatment planning; however, substantial intratumoral heterogeneity, complex multimodal MRI patterns, and limited interpretability challenge existing deep learning approaches. This study aims to develop ProtoSurv-X, an explainable and uncertainty-aware framework for MRI-based glioma survival prediction that integrates probabilistic tumor habitat modeling, prototype-guided learning, evidential prediction, and evidence-grounded clinical explanations. A unified cohort was constructed from the BraTS 2019 and BraTS 2020 datasets by removing duplicate subjects, resulting in 369 unique subjects, including 118 gross total resection patients with complete survival annotations. The proposed framework uses diffusion-enhanced SwinUNETR for tumor segmentation, probabilistic habitat construction, and fusion of radiomic, deep imaging, habitat, and age-related features. Prototype learning and survival-aware contrastive learning generate prognostic representations, while Evidential Deep Learning estimates risk and predictive uncertainty alongside continuous survival regression. For explanation generation, seven open-source large language models were benchmarked using structured model evidence, and a Meta-Llama-3.1-8B-Instruct model was fine-tuned using QLoRA. In five-fold cross-validation on the unified survival cohort, ProtoSurv-X achieved a mean absolute error of 145.2 ± 6.5 days, an RMSE of 189.4 ± 8.3 days, a C-index of 0.745 ± 0.008, and an integrated Brier score of 0.124. The segmentation module achieved Dice scores of 91.24%, 87.38%, and 81.76% for whole tumor, tumor core, and enhancing tumor, respectively, on BraTS 2019. The fine-tuned explanation model obtained a BERTScore of 0.915 and a hallucination rate of 2.6%. These findings indicate that ProtoSurv-X offers a unified computational framework for accurate, uncertainty-aware, and evidence-grounded glioma prognostic modeling, while further clinical and multicenter validation remains necessary.

Automated Retinal Disease Classification from OCT Images Using a Lightweight CNN–SE Architecture

2026-10-05T06:29:58+07:00

Automated classification of retinal diseases from Optical Coherence Tomography (OCT) images can support timely and consistent identification of retinal abnormalities; however, many existing approaches rely on computationally intensive architectures or complex preprocessing procedures. This study proposes a lightweight Convolutional Neural Network with Squeeze-and-Excitation blocks (CNN–SE) for automated multi-class retinal disease classification while maintaining a compact computational structure. The proposed architecture consists of convolutional layers with SE-based channel recalibration, global average pooling, a fully connected layer, and dropout. Unlike approaches that depend on pretrained backbones or multi-stage segmentation, the proposed framework integrates channel-wise feature recalibration directly into a compact CNN, allowing the model to learn discriminative retinal representations with limited architectural complexity. Data augmentation was applied only to the training images, and the network was optimized using the Adam optimizer. The model was evaluated on the eight-class OCT-C8 dataset and the four-class OCT2017 dataset using accuracy, precision, recall, F1-score, and macro-ROC-AUC. The proposed model achieved 97.00% accuracy on OCT-C8 and 99.00% on OCT2017. To assess domain generalization, the model was trained exclusively on OCT-C8 and directly evaluated on the independent OCT2017 test set without retraining or fine-tuning. Across five random seeds, the cross-dataset evaluation achieved 98.74 ± 0.11% accuracy, with consistently high precision, recall, F1-score, and macro-ROC-AUC. The proposed network maintains a compact computational footprint and enables efficient image inference compared with the evaluated lightweight architectures. These results demonstrate that the proposed CNN–SE architecture provides strong classification performance and cross-dataset generalization while maintaining computational efficiency, making it a promising lightweight framework for automated retinal OCT image analysis

A Dual-Stage Deep Lesion Segmentation Framework with Gradient and Boundary Optimization for Diabetic Retinopathy Retinal Images

2026-08-11T18:12:05+07:00

Diabetic Retinopathy (DR) is a leading global cause of preventable blindness, and the early detection and precise segmentation of pathologies play a key role in clinical decisions in this disease. Conventional deep learning models, however, suffer from the problem of weak lesion boundaries, gradient inconsistency, and structural distortion in heterogeneous datasets. To overcome these drawbacks, we present a sophisticated systemthat combines GEONet (Gradient and Edge-Optimized Network) to achieve accurate gradient- and edge-aware segmentation, and BESNet (Boundary-Enhanced Segmentation Network) to introduce boundary refinement and structural preservation. This two-step strategy can guarantee the high-fidelity capture of subtle retinal lesions with anatomical consistency. Comparative analysis was carried out against state-of-the-art baselines. Experimental results on the EyePACS dataset demonstrate the effectiveness of the proposed framework. The integrated GEONet+BESNet architecture achieved a Dice Coefficient of 89.0%, Intersection-over-Union (IoU) of 87.0%, Boundary Accuracy of 89.0%, and a Hausdorff Distance of 5.1, outperforming all competing methods in terms of segmentation fidelity and boundary preservation. Structural consistency was further enhanced, attaining a Dice Similarity Coefficient (DSC) of 90.5% and a Structural Similarity Index Measure (SSIM) of 90.3%. From a clinical screening perspective, the proposed framework achieved a Precision of 92.8% and Specificity of 94.6%, indicating reliable lesion localization with a reduced rate of false-positive detections. These findings confirm that the synergistic integration of gradient-aware segmentation through GEONet and boundary-enhanced refinement through BESNet effectively preserves lesion morphology and improves segmentation robustness across heterogeneous retinal imaging conditions.

A Multimodal Graph Neural Network for Multiclass ADHD and ASD Classification with Leakage-Aware Evaluation

2026-08-11T18:19:50+07:00

Neurodevelopmental disorders such as attention deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD) share overlapping clinical symptoms, complicating diagnosis and motivating objective, data-driven approaches using neuroimaging and machine learning. Graph neural networks (GNNs) have shown strong performance in this domain, yet many existing studies rely on single-modality data, transductive learning, and feature selection procedures that may introduce information leakage and inflate reported accuracy. This study proposes a multimodal graph learning framework that integrates resting-state fMRI (rs-fMRI), structural MRI (sMRI), and demographic data for classifying ADHD, ASD, and healthy controls (HC). The framework integrates temporal stability-based functional connectivity, hybrid feature selection, and adaptive multi-graph learning to exploit complementary information across these modalities. Using the ADHD-200 and ABIDE datasets, the framework is evaluated under three protocols that progressively tighten control over information leakage: transductive learning with global feature selection, inductive learning with global feature selection, and inductive learning with fold-wise feature selection. Results show that classification performance is highest under the transductive, globally-selected setting (85.5% accuracy for HC vs ADHD vs ASD, 92.5% for HC vs ASD, and 90.4% for HC vs ADHD), but decreases under the strictest leakage-aware protocol (70.9%, 81.7%, and 79.3%, respectively). This performance gap indicates that conventional evaluation protocols can substantially overestimate real-world generalization. Importantly, the proposed framework still achieves reasonable accuracy under the strictest setting, suggesting genuine discriminative capability beyond evaluation artifacts. These findings emphasize that leakage-aware evaluation, although yielding lower numbers, provides a more realistic and trustworthy estimate of model performance, highlighting its importance for developing reliable neuroimaging-based GNN models

Adaptive Sparse Cross-Scale Transformer for Computationally Efficient Brain Tumor MRI Segmentation

2026-08-14T10:59:40+07:00

Brain tumor segmentation from Magnetic Resonance Imaging (MRI) is a critical task in medical image analysis, owing to the irregular morphology of tumors, heterogeneous intensity distributions across MRI modalities, and the presence of multiple overlapping sub-regions. Accurate delineation of these regions is essential for clinical diagnosis, treatment planning, and longitudinal disease monitoring. Although Vision Transformer (ViT)-based architectures have demonstrated promising performance in medical image segmentation by capturing long-range dependencies and global contextual information, conventional multi-scale transformer models remain constrained by static grouped attention mechanisms that introduce redundant computations and exhibit quadratic complexity with respect to input size. To address these limitations, this paper proposes the Adaptive Sparse Cross-Scale Transformer (ASCT), a computationally efficient framework for brain tumor MRI segmentation. ASCT incorporates three key innovations: (i) a Dynamic Scale Routing (DSR) module that adaptively weights multi-scale features using learned routing coefficients, replacing fixed channel grouping; (ii) a Sparse Token Attention (STA) mechanism that restricts attention computation to the most informative token pairs, reducing complexity from quadratic O(N²) to near-linear O(Nk); and (iii) a linearized attention approximation that significantly reduces GPU memory consumption during training. Additionally, cross-scale feature fusion is performed prior to the attention operation to suppress redundant computations and enhance inter-scale feature interaction. The proposed ASCT model is evaluated on the BraTS 2021 dataset for multi-region segmentation, targeting Whole Tumor (WT), Tumor Core (TC), and Enhancing Tumor (ET) sub-regions. Experimental results demonstrate that ASCT achieves Dice scores of 89.2%, 84.1%, and 83.5% for WT, TC, and ET, respectively, yielding an average Dice score of 85.6%. Compared to the baseline Vision Transformer, the proposed model reduces computational complexity from 94.6 GFLOPs to 58.4 GFLOPs and GPU memory usage from 11.2 GB to 7.3 GB, confirming its efficiency and practical viability for real-world clinical applications.

Structured Nursing Handover Report Generation from Clinical Speech using Fine-Tuned XLSR-53 and T5: A Benchmarking Study

2026-08-15T22:30:44+07:00

Accurate nursing handovers are critical for patient safety, as miscommunication during shift transitions leads to irreversible clinical errors. This work proposed an end-to-end pipeline that converts unstructured clinical nursing speech into standardized handover reports using a fine-tuned XLSR-53 acoustic model and T5-base text-to-text transformer. An Australian English clinical corpus of 200 synthetic nursing handover recordings from the CSIRO data access portal was utilised in this work. This benchmarking study was conducted within the CSIRO synthetic Australian English nursing handover corpus and does not represent a broad cross-domain clinical ASR benchmark. A domain-specific benchmarking study across seven state-of-the-art ASR architectures (Whisper Tiny/Base/Small, Wav2Vec2 Base/Large, HuBERT Large, XLSR-53) was conducted using this corpus. The experimental results further revealed XLSR-53 as the optimal architecture for clinical nursing speech recognition. A partial layer-freeze strategy was adopted in XLSR-53 by freezing the first 12 of 24 encoder layers, empirically validated through an ablation study with five freeze configurations (L=0, 6, 12, 18, 24). XLSR-53 preserves cross-lingual phonetic representations while enabling clinical vocabulary adaptation. A clinically motivated evaluation framework using curated medical vocabulary terms computes Medical Precision, Recall, and F1-Score along with standard WER, CER, and PER to assess reliability in clinical term recognition. Benchmarking against Google Health AI's MedASR zero-shot revealed that the proposed system XLSR-53 (L=12) achieved 17.15% WER against MedASR's 28.87% (p<0.001) with a Medical F1-Score of 0.98 and ROUGE-L of 0.92. Although results were obtained on synthetic Australian English speech, performance under real clinical conditions with background noise, overlapping speakers, and spontaneous interruptions requires further validation.

Adaptive Optimizers for Neural Network Sperm Motility Classification

2026-10-03T17:10:00+07:00

Infertility remains a major challenge in global reproductive healthcare, with abnormal sperm motility being one of the primary contributing factors to male infertility. Conventional sperm motility analysis relies heavily on manual microscopic observation, which is subjective, time-consuming, and prone to inter-operator variability. Although Computer-Aided Sperm Analysis (CASA) systems provide automated solutions, their high cost and limited accessibility restrict widespread clinical adoption. Therefore, there is a need for a lightweight and cost-effective computational approach that can automatically classify sperm motility abnormalities with high accuracy. This study proposes a machine learning framework based on a Multilayer Perceptron (MLP) neural network optimized using adaptive optimization algorithms to classify sperm motility trajectories. The model utilizes two kinematic features extracted from microscopic video tracking: mean velocity and trajectory linearity. A dataset consisting of 276 unique sperm trajectories was obtained from microscopic recordings and processed using the Trackpy library to generate motion trajectories. The dataset was divided using an 80:20 train–test split strategy to evaluate the generalization capability of the model. Four adaptive optimization algorithms were systematically evaluated, including Adagrad, Adadelta, FTRL, and Adamax. The neural network was trained for 50 epochs using binary cross-entropy as the loss function and evaluated using accuracy, per-class precision, recall, F1-score, specificity, balanced accuracy, and ROC-AUC. Experimental results demonstrate that the Adamax optimizer achieved the best performance with an accuracy of 96.43% and a perfect ROC-AUC of 1.00, outperforming Adagrad (92.86%), FTRL (69.64%), and Adadelta (67.86%). The results indicate that Adamax provides more stable convergence and better optimization behavior for dense kinematic feature spaces. These findings highlight the effectiveness of adaptive optimization techniques in improving neural network performance for automated sperm motility classification. The proposed approach offers a computationally efficient and accessible alternative for supporting clinical sperm analysis and reducing subjectivity in reproductive diagnostics

Adaptive Swarm-Driven Skin Lesion Segmentation and Classification Using Boundary-Aware U-Net and Hybrid Vision Transformer

2026-10-06T08:10:51+07:00

Skin cancer is one of the most prevalent dermatological diseases worldwide, and timely and accurate diagnosis is essential for effective treatment and improved patient outcomes. However, automated skin lesion analysis remains challenging due to irregular lesion boundaries, low contrast between lesions and surrounding skin, imaging artifacts, variations in lesion appearance, and significant intra-class variability. To address these challenges, this research proposes an Adaptive Swarm-Driven Skin Lesion Segmentation and Classification Approach that integrates an Adaptive Boundary-Aware U-Net (ABAU-Net), a Butterfly Optimization Algorithm–Elephant Herding Optimization (BOA-EHO) strategy, and a Hybrid Vision Transformer with Dynamic Feature Fusion (HViT-DFF). The proposed ABAU-Net is designed to enhance segmentation accuracy by emphasizing boundary information and preserving fine-grained lesion structures, particularly in cases involving poorly defined or irregular borders. The BOA-EHO optimization strategy combines global exploration and local exploitation mechanisms to improve network parameter optimization, accelerate convergence, and reduce the possibility of premature convergence. Following segmentation, the HViT-DFF model integrates CNN-based local spatial features with transformer-based global contextual representations through dynamic feature fusion, enabling robust discrimination among multiple skin lesion classes. Extensive experiments conducted on the ISIC skin lesion dataset demonstrate the effectiveness of the proposed framework, achieving a Dice coefficient of 98.2%, Jaccard index of 97.5%, classification accuracy of 98.8%, and AUC of 0.992. These results outperform recent CNN-, Transformer-, and hybrid-based approaches. Overall, the proposed framework demonstrates strong potential for accurate, reliable, and automated skin lesion segmentation and multi-class classification, supporting computer-aided dermatological diagnosis and potentially facilitating earlier clinical intervention.

NeuroQFormer-Net: A Quaternion-Guided Hybrid Attention Transformer Fusion Network for Accurate Brain Tumor Segmentation

2026-10-06T08:15:40+07:00

Brain tumor segmentation is a critical task in medical image analysis, as accurate identification and delineation of tumor regions support timely diagnosis, treatment planning, therapeutic decision-making, and disease monitoring. Magnetic Resonance Imaging (MRI) provides multiple complementary modalities that offer valuable information about tumor structure and tissue characteristics. However, accurate segmentation remains challenging due to substantial variations in tumor size, shape, texture, and location, along with irregular and overlapping tumor boundaries, low-contrast regions, and considerable differences in the information captured by different MRI modalities. To address these challenges, this study proposes NeuroQFormer-Net (Quaternion-Guided Hybrid Attention Transformer Feature Fusion Network), a novel deep learning framework for multi-modal brain tumor segmentation. The proposed framework employs quaternion-based inter-channel correlation learning to effectively capture and integrate complementary information across multiple MRI modalities while preserving structural dependencies and enhancing cross-modal feature representations. In addition, a hybrid attention mechanism integrated with transformer-based learning is utilized to simultaneously capture fine-grained local structural characteristics and long-range contextual dependencies. To further improve segmentation accuracy, an adaptive co-learning feature fusion strategy is introduced to effectively integrate multi-scale representations obtained from different network levels. Furthermore, a boundary-aware refinement module enhances tumor boundary localization and reduces segmentation errors, particularly in low-contrast, irregular, and complex tumor regions. Experimental evaluation demonstrates the effectiveness of NeuroQFormer-Net, achieving an average Dice Similarity Coefficient (DSC) of 96.45%, Intersection over Union (IoU) of 93.26%, Precision of 96.17%, Recall of 96.08%, and Hausdorff Distance (HD) of 2.31 mm. These results indicate that the proposed framework provides accurate and robust multi-modal brain tumor segmentation