International Journal of Biomedical Engineering and Technology (9 papers in press)

Regular Issues

• Real-Time Emotion Detection From Integrating Electroencephalography, Facial Expressions, and Speech : Review  
  by Aaditi More, Joydeep Sengupta 
  Abstract: Emotion recognition systems have gained substantial focus due to their pivotal role in man-machine interaction and affective computing applications. This comprehensive literature survey explores the latest advancements in the field, spanning a different variety of strategies and datasets. Survey delves into realm of cross-corpus speech emotion identification, discussing innovative approaches including deep local domain adaptation and multimodal systems like RobinNet. Furthermore, it investigates electroencephalography-based emotion recognition techniques, highlighting hierarchical self-attention networks, deep forest models, and spatio-temporal convolution attention neural networks. The paper also presents collaborative frameworks for the diagnosis of sadness that makes use of cross-scale facial feature analysis and negative emotion detection. In realm of machine learning, ensemble approaches for affective computing and the efficacy of prompt consistency in multi-label textual emotion detection are examined. Through this survey, emerging trends, comparative studies, and validation frameworks in emotion recognition systems research are synthesised. The findings underscore the significance of these systems in knowing human emotions and creating the groundwork for next developments in affective computing.
  Keywords: Speech Emotion Recognition; Electroencephalography Based Emotion Recognition; Deep Learning Models; Multimodal Emotion Recognition.
  DOI: 10.1504/IJBET.2025.10076819
   
  
• MA-TIL-GBNN: A Multi-Attention Enhanced Deep Incremental Learning Based Diabetes Prediction and Drug Recommendation Framework  
  by Netra S. Patil, Naveenkumar Jayakumar, Rohini B. Jadhav, Gauri R. Rao, Shashank D. Joshi, Shubhangi R. Katkar, Madhavi Mane 
  Abstract: This research is designed to develop an advanced framework for the prediction of diabetes and the personalized recommendation of drugs. This research proposes a Multi-Attention enhanced Task Incremental learning coupled Gradient Boost Deep Neural Network (MA-TIL-GBNN) approach aimed at the prediction of diabetes and recommendation of drugs based on their types.The approach integrates MA techniques into the DNN and Generative Adversarial Network-based data augmentation (GDA) to effectively analyse the health indicators. Additionally, the TIL enhances the training of data and the Light GBM facilitates efficient processing. The experimental results using Diabetes Health Indicators Dataset showed 97.79% accuracy, 98.22% sensitivity, and 97.36% specificity respectively.
  Keywords: Diabetes prediction; Drug Recommendation; Deep Learning; Machine Learning; Healthcare Management.
  DOI: 10.1504/IJBET.2026.10077448
   
  
• Simulation and design of an elliptical surface coil for small animal MRI at 3T  
  by Giulio Giovannetti, Benjamin Michael Hardy, Francesca Frijia, Alessandra Flori, Vincenzo Positano 
  Abstract: Custom-designed radiofrequency coils are commonly utilised in preclinical magnetic resonance imaging (MRI) to image small animals due to their cost-effectiveness and flexibility in adaptation to specific anatomical regions. Researchers frequently prefer such specialised coils for targeted metric assessment because their handcrafted nature allows for precise customisation. Rather than repeated experimental iterations, simulation-based refinement of coil architecture streamlines the design process. Numerical simulation methods offer more accurate estimations of signal-to-noise ratio (SNR) compared to magnetostatic models. The present study presents a comprehensive validation, using finite-difference time-domain (FDTD) full-wave simulation, of an elliptical radiofrequency (RF) coil tailored for small animal MRI applications. The approach incorporates calculations of coil and sample-induced resistances, inductance parameters, and magnetic field distributions under loading conditions with both phantom and whole-body mouse models. The accuracy of the simulation data is verified with data acquired from a transmit/receive elliptical coil prototype for a 3T MRI clinical scanner.
  Keywords: magnetic resonance; radiofrequency coils; inductance; magnetic field; resistance; signal-to-noise ratio; SNR.
  DOI: 10.1504/IJBET.2026.10078039
   
  
• Piezoelectricity in biomedical innovation: a systematic review of human-centric devices, applications, challenges and future directions  
  by Fatima Hassan, Taha Sana, Hamna Rana 
  Abstract: Piezoelectricity has emerged as a key mechanism in biomedical engineering, enabling localised electrical stimulation, sensing, and energy harvesting in human-centric devices. This systematic review analyses recent advances (2021-2025) in piezoelectric materials, device architectures, and biomedical applications, including tissue engineering, implantable and wearable systems, biosensors, neural interfaces, and drug delivery. Polymer-based materials such as PVDF exhibit superior flexibility and biocompatibility, whereas ceramic materials provide higher electromechanical efficiency but face limitations related to toxicity and mechanical mismatch. Despite their potential for self-powered operation and bioelectric modulation, clinical translation remains constrained by low power output, signal-to-noise limitations, material instability, and integration challenges with biological tissues. This review identifies material innovation, device miniaturisation, and system integration as key barriers to deployment, and highlights future directions toward lead-free nanomaterials, flexible hybrid electronics, and scalable biomedical applications.
  Keywords: piezoelectric biomaterials; implantable biomedical devices; bioelectric stimulation; self-powered sensing; flexible piezoelectric systems; clinical translation.
  DOI: 10.1504/IJBET.2026.10078103
   
  
• Enhanced slicing adversarial network with attention and multi-resolution generators for high-fidelity MRI reconstruction  
  by Libya Thomas, Joseph Zacharias 
  Abstract: We introduce SAN++, an enhanced slicing adversarial network for compressed-sensing MRI reconstruction that integrates three key novelties: a transformer-guided attention block (TGAB), an edge-aware adaptive sampling module (EAASM), and a self-supervised pretraining strategy using masked image modelling (MIM). SAN++ extends prior GAN-based MRI methods by incorporating global transformer-based attention to better capture long-range dependencies, and by learning dynamic k-space sampling masks guided by salient image edges, which preserves critical structural features. We pretrain the network with a masked reconstruction task on large unlabeled datasets, then fine-tune adversarially with multi-resolution generators and a sliced optimal transport loss. Experiments on MRI datasets under various undersampling ratios (4x, 8x) and noise levels demonstrate that SAN++ outperforms DAGAN, RefineGAN. SAN++ achieves higher PSNR/SSIM and lower LPIPS perceptual error across settings, and shows robust performance under noise and sampling variability. An ablation study confirms each components benefit, notably a 23 dB PSNR gain from TGAB and EAASM. Our results (with quantitative tables and example reconstructions) highlight the efficacy of combining transformer-guided attention, adaptive sampling, and self-supervised pretraining in adversarial MRI reconstruction.
  Keywords: CS-MRI; undersampling pattern; trade-off; neural network modified GAN.
  DOI: 10.1504/IJBET.2026.10078230
   
  
• Leakage-Safe Case-Level Evaluation of ResNet50V2 for Benign-Malignant Liver Tumour Classification on CT Slices  
  by Smitha B, Vinod Kumar, Kumar S. S 
  Abstract: Binary (benign vs. malignant) liver tumour classification from computed tomography (CT) slices is often evaluated using slice-level splits that can introduce patient-level leakage. We present a leakage-safe, case-level evaluation of ResNet50V2 using stratified 5-fold cross-validation repeated over three random seeds. Slice probabilities were aggregated per Case_ID using mean probability (primary) and vote rate (secondary), with decision thresholds derived only from validation folds. Two preprocessing variants were assessed (BASIC and HE_GAUSS_CONTRAST). Across 15 runs, pooled case-level discrimination was AUC = 0.765 +_
  Keywords: leakage-safe evaluation; case-level cross-validation; liver tumour classification; computed tomography (CT); ResNet50V2; slice aggregation; calibration.
  DOI: 10.1504/IJBET.2026.10078321
   
  
• Structured Radiology Report Generation Using ViT-B/16 and Clinical-T5: A Multimodal Approach on IU X-Ray Dataset  
  by Nilam Khairnar, Shirish Sane 
  Abstract: This work presents a multimodal framework for generating structured radiology reports directly from the chest X-ray images. The model combines the Vision Transformer (ViT-B/16) to extract rich visual representations with Clinical-T5, which is a domain-tuned language model that interprets the clinical text using a co-attention module. The visual and textual streams interact to allow the system to link features of images with clinical descriptions that match the image. The decoder then generates well-structured reports containing Indication, Findings, and Impression sections. The framework was trained and evaluated on the IU X-Ray dataset and was found to have significant advances in the BLEU and ROUGE-L metrics, and strong CheXbert F1 results compared with previous methods. The results show that a combination of transformer-based vision and language models can generate coherent, interpretable, and clinically reliable radiology reports, highlighting the importance of multimodal learning for automated radiology report generation.
  Keywords: BLEU; Clinical T5; NLP; Radiology Report Generation; ROUGE; Vision Transformer.
  DOI: 10.1504/IJBET.2026.10078324
   
  
• Experimental Analysis on Effect of Type-2 Diabetic Mellitus: on Strength Behaviour of Trabecular Bone Structures  
  by Swapnil S. Barekar, Tushar A. Jadhav, Navin Kumar 
  Abstract: The type-2 diabetes affected bones are more susceptible to fragile fractures due to the deterioration in bone mineral density (BMD). This study aims to investigate the structural integrity of human trabecular bone by comparing nanoindentation-based mechanical evaluation outcomes of three diabetic and two non-diabetic bone samples. The obtained results stated that a 45-year-old T2DM sample has lower modulus of elasticity (40.81%), toughness (34.24%) and diminished bone quality when compared to normal bones. Mechanical analysis shows that T2DM causes mean reductions of the elastic modulus by 5.3 GPa, stiffness to 4.9 N/nm, and hardness by 0.27 GPa in bones. The reverse problem-solving method also effectively approximates yield stress, yielding results comparable to direct calculations that eliminate the requirement for another destructive testing. The findings also highlight the significance of multi-parametric analysis in fracture-risk assessment strategies, other than focusing only on bone mineral density for individuals with T2DM.
  Keywords: Bone Defects; Trabecular Bone; T2DM; Berkovich Nano Indentation; Elastic Modulus; Yield Strength.
  DOI: 10.1504/IJBET.2026.10078458
   
  
• Predicting Adolescent Type 2 Diabetes Using an Intelligent System  
  by Mrko Arsenovic, Juliens Baker, Livija Cveticanin 
  Abstract: Type 2 diabetes mellitus (T2D) in adolescents is a rising public health concern, yet early detection is challenged by limited and imbalanced datasets. This study compares traditional machine learning (ML) and deep learning (DL) methods for predicting adolescent prediabetes using a dataset of 209 records (11.5% prediabetic). Preprocessing included imputation, feature reduction, normalisation, and SMOTE-based class balancing. Baseline ML models logistic regression, random forest, support vector machine (SVM), and multi-layer perceptron were evaluated via stratified cross-validation, with SVM achieving the best F1-score (0.955). To address class imbalance, synthetic data were generated using generative adversarial networks (GANs) and Wasserstein GANs (WGANs). A 1D convolutional neural network (CNN) trained on the augmented dataset achieved 98.5% accuracy, 96.0% precision, 97.0% recall, and a 96.5% F1-score on the original test set. Results confirm the value of GAN-based augmentation combined with CNNs for improving prediction under limited data, supporting timely T2D risk identification in adolescents. Unlike previous machine learning studies that relied solely on statistical resampling or small neural models, our approach combines GAN-based data synthesis with a one-dimensional convolutional network, yielding a substantial improvement in predictive power and generalisability under limited data conditions.
  Keywords: type 2 Diabetes Mellitus in adolescents; predictive modeling; Convolutional Neural Network (CNN); Generative Adversarial Networks (GAN); data augmentation; class imbalance; medical decision support.
  DOI: 10.1504/IJBET.2026.10078672